43 results about "Data driven algorithms" patented technology

The invention discloses a coal drop pipe full-life-cycle operation and maintenance decision-making method. The method comprises the steps of S1, coal drop pipe full-life-cycle data collection; s2, constructing and updating a digital twinborn body of the coal drop pipe: S2.1, constructing a multi-dimensional digital twinborn body comprising a geometric model, a physical model, a behavior model and a rule model, and comprehensively describing the characteristics of each aspect of the coal drop pipe; s2.2, dynamically updating the twinborn body, and dynamically updating the twinborn body through a data driving algorithm by using real-time monitoring data; s3, health state evaluation and reliability prediction, including short-term health diagnosis and long-term reliability prediction; s4, carrying out dynamic operation and maintenance decision making based on risks and cost; and S5, performing decision issuing execution and closed-loop feedback. According to the method, the design data, the real-time monitoring data, the environment data and the operation and maintenance historical data are integrated, the coal drop pipe digital twinborn body is constructed and dynamically updated, it is ensured that the model is consistent with the physical entity state, and the comprehensiveness and timeliness of the data are improved.

The invention discloses an intelligent fault diagnosis early warning system for mine electromechanical equipment, which comprises a data sensing module used for collecting multi-mode state data of the mine electromechanical equipment and carrying out anti-interference preprocessing, and an edge intelligent module used for receiving the preprocessed data and executing fault diagnosis and dynamic threshold adjustment. The cloud evolution module is used for receiving a diagnosis result and original data and completing self-evolution through data enhancement and model updating, and the application service module provides graded early warning, residual life prediction and operation and maintenance decisions based on the diagnosis result and evolution information; the method effectively captures early weak fault features through multi-modal data deep fusion in combination with a physical mechanism model and a data-driven algorithm, greatly improves the recognition accuracy of major faults, and improves the recognition accuracy of the major faults through deploying a lightweight CNN-LSTM model at an edge end and focusing on model evolution and data enhancement at a cloud end. And collaboration of edge quick response and cloud continuous optimization is realized.

The invention relates to the field of nonlinear system modeling, and particularly provides a global linear modeling method and system for a nonlinear system based on Lie derivative and sparse recognition, a terminal and a medium, and the method comprises the steps: obtaining an explicit nonlinear kinetic equation of a target system, calculating the Lie derivative of a state variable or an output variable of the system based on the kinetic equation, constructing a candidate observation function library directly associated with the physical mechanism of the system; performing screening and dimension reduction on the candidate observation function library by adopting a sparse recognition method to obtain a low-dimensional observation function set; and on the basis of the screened observation function and system operation data, a finite-dimensional global linear model of the target system in the dimension raising observation space is obtained through identification by using a data driving algorithm. According to the method, the online solving efficiency and the control real-time performance are improved, and efficient and reliable multi-target collaborative optimization control of systems such as a wind driven generator is realized.

This invention proposes a data-driven back-end parameter identification method for underwater wireless power transfer systems, belonging to the fields of wireless power transfer technology and power electronics technology. By employing a data-driven method based on the XGBoost data-driven algorithm, a nonlinear complex mapping model between voltage parameters and coupling parameters is established. At the back end, the multivariate parameters to be identified in the coupling mechanism are inverted based on the Buck input voltage. This method only requires DC voltage acquisition to achieve online identification of coupling parameters, eliminating the need for bilateral communication modules or high-frequency sampling modules, thus reducing hardware costs. It enables rapid identification of multivariate coupling parameters in underwater magnetically coupled wireless power supply systems, exhibiting better adaptability to marine environments. This method solves the problems of existing underwater wireless power transfer systems relying on complex analytical models and requiring bilateral communication and acquisition for parameter identification, while also addressing the difficulty of rapid online identification of coupling parameters in complex underwater environments such as seawater.

The invention discloses a digital intelligent PHM platform for a driving mechanism of coal unloading and storage mechanical equipment of a coal wharf, which comprises five core parts, namely a data acquisition layer, a data transmission layer, a data storage and processing layer, a digital modeling layer and a visualization and decision support layer, and is characterized in that the data acquisition layer comprises a multi-modal sensor arranged at a key part of the equipment; the data storage and processing layer comprises a framework combining edge computing and cloud computing, performs cleaning, noise reduction and feature extraction on original data, and constructs a driving mechanism operation state portrait through a multi-source information fusion technology; the digital modeling layer comprises fusion based on physical modeling and a data driving algorithm, and equipment digital twin bodies capable of being updated in real time are formed; the visualization and decision support layer provides a visual interface and operation and maintenance strategy recommendation; according to the invention, by realizing unified data acquisition and centralized management, the problems of data dispersion and incompatibility in a traditional monitoring system are solved, and a data foundation is laid for full-life-cycle management of equipment.

The invention provides an equipment attitude identification-based wireless charging system anti-offset control method, which comprises the following steps of: firstly, acquiring electrical information of a primary side system under different roll angles and butt joint distances of equipment, specifically voltage and current at two ends of a shunt capacitor and a phase difference between the voltage and the current; the method comprises the following steps: establishing an equipment docking attitude parameter identification model by taking voltage and current at two ends of a shunt capacitor and a phase difference between the voltage and the current as input information and taking a roll angle and a docking distance of equipment as output information, and identifying a docking attitude of the equipment before starting stable charging to ensure that the equipment enters a charging process in an optimal attitude; in the stable charging process, the posture of the equipment is continuously monitored, and the problems of unstable charging, efficiency reduction and the like caused by equipment deviation due to factors such as ocean current impact and the like are solved in real time.

This invention discloses a model for estimating the temperature of coolant in the main pipeline of a nuclear reactor, considering measurement errors. Belonging to the field of nuclear power plant temperature measurement technology, the model includes the following steps: acquiring actual temperature measurement data from the power plant and establishing a dataset corresponding to the measured data and agreed-upon true values; analyzing the actual temperature measurement process and classifying measurement errors; considering different error distribution characteristics, combining a data-driven algorithm with a stochastic model to establish a temperature estimation model; designing a solution method for estimating prior information and calculating model coefficients in the temperature estimation model; optimizing and solving the temperature estimation model using the created dataset, and verifying its accuracy and reliability. This invention solves the problems of difficulty in detailed study of existing measurement errors and difficulty in solving for the true distribution of coolant temperature in nuclear pipelines.

The invention provides a blade root load-based wind power mixed tower load estimation method, which comprises the following steps of: firstly, building a wind turbine generator multi-body dynamics simulation model, simulating dynamic responses under different operation conditions, and generating a simulation data set; then constructing a load estimation model of a fusion mechanism and a data-driven algorithm, and training and verifying the load estimation model by using the simulation data set; and acquiring real-time operation state parameters and blade root loads of the wind turbine generator in an actual wind turbine generator, inputting real-time data into the trained load estimation model, and outputting load estimation values of all sections of the mixed tower in real time as a basis for structural health assessment and control optimization. Finally, a blade root load sensor data validity monitoring and fault-tolerant switching mechanism is introduced, and continuity and reliability of the load estimation process are guaranteed. The method is high in engineering adaptability, easy and convenient to deploy and suitable for various practical application scenes such as structural health monitoring, fatigue life evaluation and intelligent control strategy optimization of the wind turbine generator.

The invention provides a garment simulation method and system based on simulated garments and data driving and application. The method comprises the steps that the simulated garments and garment model grids of the simulated garments are acquired; preprocessing is carried out on the basis of a garment process related to the simulation garment, and animation data of the simulation garment in different postures are generated; smoothing the animation data to generate a garment skeleton; inputting the current human body posture data into a pre-trained prediction network model to predict transformation information of each skeleton on the garment skeleton; and adjusting the clothing skeleton based on the transformation information, and further driving the clothing model grid through the adjusted clothing skeleton. According to the scheme, high-real-time, high-quality and low-resource-consumption garment simulation can be realized, computing resources can be effectively utilized, and the complexity of calculation is reduced, so that a high-quality garment effect is realized on wider equipment. And a data-driven algorithm and a machine learning model are adopted to solve the clothing effect in real time, so that the clothing can adapt to the action of the user more quickly.

The invention discloses a wind turbine generator tower drum monitoring system with a self-healing function. The system comprises a sensing layer, an analysis layer, a decision-making layer and a self-healing layer. The sensing layer collects multi-source response data of the wind turbine generator tower drum in real time, and full-system holographic sensing of the wind turbine generator tower drum is achieved; the analysis layer combines a multi-source data fusion algorithm and a data driving algorithm to carry out real-time state monitoring and future state prediction on the wind turbine generator tower drum prestressed cable; a decision-making layer makes a next action strategy of the fan according to analysis results of a multi-source data fusion algorithm (present) and a data driving algorithm (future) and a linkage decision of the main control system; and the self-healing layer actively triggers the repair system to execute targeted repair operation in real time, so that self-adaptive recovery of the prestressed cable is realized, and stable operation of the wind turbine generator is ensured. According to the invention, a monitoring and self-healing integrated intelligent solution can be provided for operation and maintenance of the tower drum of the wind turbine generator.

This invention discloses a method for predicting the three-dimensional spatiotemporal distribution of algae in aquatic bodies based on a physical information neural network (PINN), belonging to the field of algal bloom prediction and control technology. Addressing the problems of poor generalization of purely data-driven models, difficulty in fitting purely mechanistic models, and insufficient vertical fitting ability in existing algal distribution prediction methods, this invention integrates the advantages of physical mechanisms and data-driven algorithms. It embeds the advection-diffusion-response (ADR) equation of algal growth into the loss function of a physical information neural network (PINN), constructing a hybrid prediction model that combines data fitting accuracy with physical constraints. This achieves accurate prediction of the three-dimensional spatiotemporal distribution of algae and inversely obtains physically meaningful ecological parameters. This invention improves the accuracy, generalization, and physical interpretability of algal distribution prediction and can be effectively applied to the monitoring, early warning, and aquatic ecological environment management of harmful algal blooms.

The invention provides an AI monitoring method for analyzing the corrosion risk of a drilling pipeline based on time sequence data. The AI monitoring method comprises the following steps: S1, acquiring multi-source time sequence data of the drilling pipeline of an offshore drilling platform; s2, preprocessing and standardizing the original data; s3, time series data feature engineering and feature screening; s4, marking corrosion risk labels and dividing a data set; s5, training an AI corrosion risk prediction model; s6, verifying and optimizing the performance of the AI model; s7, performing real-time corrosion risk reasoning; s8, carrying out corrosion risk early warning and data feedback; and S9, historical data redisk and model iterative optimization are carried out. Aiming at the corrosion risk of the drilling pipeline of the offshore drilling platform, a full-process AI monitoring system of data driving, algorithm fusion and closed-loop iteration is constructed, and aims to realize real-time accurate identification, graded early warning and continuous optimization of the corrosion risk, guarantee safe operation and maintenance of the platform and reduce the operation and maintenance cost.

The application relates to the field of nonlinear system modeling, and specifically provides a nonlinear system global linearization modeling method, system, terminal and medium based on Lie derivatives and sparse identification, the method comprising the following steps: obtaining an explicit nonlinear dynamic equation of a target system; based on the dynamic equation, constructing a candidate observation function library directly associated with the physical mechanism of the system by calculating Lie derivatives of system state variables or output variables; screening and reducing the dimension of the candidate observation function library by using a sparse identification method to obtain a low-dimensional observation function set; and based on the screened observation function and system operation data, identifying a limited-dimensional global linear model of the target system in the dimensioned observation space by using a data-driven algorithm. The application improves online solving efficiency and control real-time performance, and realizes efficient and reliable multi-target collaborative optimization control of systems such as wind driven generators.

This invention provides an AI monitoring method for drilling pipeline corrosion risk based on time-series data analysis, comprising the following steps: S1: Multi-source time-series data acquisition for offshore drilling platform drilling pipelines; S2: Raw data preprocessing and standardization; S3: Time-series data feature engineering and feature selection; S4: Corrosion risk labeling and dataset partitioning; S5: AI corrosion risk prediction model training; S6: AI model performance verification and optimization; S7: Real-time corrosion risk inference; S8: Corrosion risk early warning and data feedback; S9: Historical data review and model iterative optimization. This invention addresses the corrosion risk of offshore drilling platform drilling pipelines by constructing a full-process AI monitoring system of "data-driven - algorithm fusion - closed-loop iteration," aiming to achieve real-time and accurate identification, hierarchical early warning, and continuous optimization of corrosion risks, ensuring safe platform operation and maintenance while reducing maintenance costs.

The invention belongs to the technical field of crossing of energy storage systems, data processing and data driving algorithms, and discloses an energy storage system single battery consistency detection method based on density peak clustering and evidence classification, a medium and a program product. Analyzing the association relationship between the historical data of the energy storage system and the related influence factors, and constructing a historical sample library; establishing an energy storage system feature extraction model based on a clustering algorithm, and dividing each battery monomer of the energy storage system into a plurality of categories; and carrying out decreasing mapping on the distance between the real-time operation data and the class center to obtain similarity, completing evidence distribution, calculating and identifying framework evidence, and judging and positioning inconsistent monomers. According to the invention, the accuracy and reliability of consistency detection of the energy storage system can be improved, inconsistent battery monomers in the energy storage system can be accurately positioned, and the reliability and safety of operation of the energy storage system can be improved.

The invention provides a gas inspection path optimization method and system based on a GPS, and relates to the technical field of gas inspection path optimization. The method comprises the steps of collecting core information; constructing a constraint inspection road network; performing routing inspection path planning to obtain an initial routing inspection path set; dynamically adjusting the inspection path to obtain an optimized inspection path set; and screening an optimal inspection scheme through a vine growth optimization algorithm. According to the method, through core logic of data driving, algorithm optimization and dynamic adaptation, gas inspection is converted from experience driving to scientific quantification, the inspection safety and efficiency are improved, meanwhile, resource consumption is reduced, and a technical scheme which can be landed and popularized is provided for safe inspection of urban gas pipelines.

The invention discloses a financial data causal relationship mining method and system for intelligent auditing. The method comprises the steps of obtaining and preprocessing structured financial data and unstructured text data; fusing domain knowledge and a data-driven algorithm to construct a financial causal graph; using a large language model to identify and quantify management layer intervention variables; constructing a structured causal model and executing anti-fact deduction to obtain a financial index prediction value under a non-intervention scene; and comparing the actual observed value with the predicted value, quantifying the manipulation effect and generating a risk clue. The invention further discloses a corresponding system which comprises a data acquisition and preprocessing function module, a financial causal diagram construction function module, a management layer intervention variable recognition and quantification function module, an anti-fact deduction function module, a manipulation effect mining function module and the like. According to the method, analysis from abnormality discovery to causal attribution can be realized, and the net influence of management layer decisions on financial results is quantified, so that the recognition capability and auditing efficiency of complex profit manipulation behaviors are improved.

The invention relates to a power battery system, in particular to a method for estimating available charge and discharge power of the power battery system. The invention discloses a method for estimating available charge and discharge power of a power battery system. The method comprises the following steps of: testing the maximum static limit charge current and the maximum static limit discharge current of a power battery under various charge states and temperatures; calculating the current average electromotive force of the battery power system; establishing a relation between a dynamic limiting current and a static limiting current based on a first-order RC equivalent circuit model; estimating the dynamic limiting current of the single battery on line in real time; comparing the dynamic limiting current of the single power battery with the maximum current allowed by the system to obtain the available current of the system, and calculating the available current and power of the system. According to the method for estimating the available charging and discharging power of the power battery system, physical modeling and a data driving algorithm are fused, and high-precision and real-time response to battery aging and complex working conditions can be achieved while the calculation efficiency is guaranteed.

This invention belongs to the field of transmission line structural safety assessment and intelligent analysis technology, and discloses a rapid assessment method for the load-bearing safety of power fittings in complex icing and wind environments. This method constructs a refined finite element analysis model including towers, transmission lines, and power fittings, introducing wind loads, icing loads, and de-icing impact loads to analyze the stress distribution and deformation characteristics of power fittings under complex icing and wind environments, identifying weak points in the fittings. It uses the Latin hypercube sampling method to design complex environmental calculation case samples, conducts large-scale finite element simulation analysis, and constructs data samples of the load-bearing characteristics of power fittings in complex icing and wind environments. Based on this, it uses data-driven algorithms such as deep neural networks, support vector machines, or random forests to establish a load-bearing safety assessment model, achieving rapid assessment of the load-bearing safety of power fittings. This invention balances assessment accuracy and efficiency, and is suitable for rapid analysis and engineering applications of the load-bearing safety of power fittings under complex icing and wind environments.

This invention discloses a mechanism-data hybrid-driven method for modeling and controlling wastewater biochemical treatment processes, comprising the following steps: establishing a mechanism model for the target pollutant; screening key kinetic parameters with high sensitivity and high volatility; using experimental datasets to invert the key kinetic parameter dataset based on the mechanism model; constructing a prediction model for the key kinetic parameters using a data-driven algorithm; coupling the prediction model and the mechanism model into a hybrid prediction model, and inputting real-time influent and operating conditions to dynamically predict the concentrations of the target pollutant and its products; based on the prediction results, solving a multi-objective optimization problem with target pollutant removal and product resource recovery as optimization objectives, and obtaining and executing the optimal control parameters. This invention reduces prediction fluctuation bias and improves prediction robustness by constraining the main mechanism framework, and improves prediction accuracy by dynamically fitting highly sensitive and highly volatile parameters using a data-driven algorithm, thereby achieving synergistic optimization of pollutant removal and resource recovery.

The application relates to the technical field of unmanned aerial vehicle scheduling and intelligent optimization, and discloses a multi-unmanned aerial vehicle cooperative task allocation method and device for complex task scenarios, which has the technical scheme as follows: global multi-source data acquisition and space-time heterogeneous hypergraph representation, complex task causal emergence decoupling and counterfactual deduction optimization, hierarchical federated heterogeneous matching degree accurate representation, distributed external generalization causal element reinforcement learning dynamic allocation decision, distributed conflict resolution and digital twin closed loop iterative optimization; by adopting frontier technologies such as space-time heterogeneous hypergraph representation, causal emergence reasoning, hierarchical federated learning, causal element reinforcement learning and digital twin closed loop optimization, a causal driving full-link algorithm architecture is constructed to break through the inherent limitations of traditional correlation-based data-driven algorithms, so that efficient cooperative operation of a large-scale heterogeneous unmanned aerial vehicle cluster in a strong coupling task and an extreme dynamic environment can be realized.

The invention discloses a DCS-based continuous flow chemical industry production multi-parameter cooperative control method, and particularly relates to the technical field of industrial process control, comprising the following steps: collecting multi-dimensional parameters in a continuous flow chemical industry production process in real time through a distributed sensor group of a DCS system; the multi-dimensional parameters comprise key parameters for directly determining a reaction state and auxiliary parameters for reflecting an operation environment. According to the method, the dynamic coupling model fusing the chemical reaction mechanism and the data driving algorithm is constructed, the complex incidence relation between key parameters can be accurately quantified and predicted, the double-closed-loop cooperative control architecture is combined, effective division and cooperation of inner-loop rapid response and outer-loop global optimization are achieved, and the method has the advantages of being high in reliability and high in reliability. The problems of multi-parameter strong coupling and control lag are effectively solved, the limitation of a traditional single-loop PID control or simple control strategy is broken through, and it is ensured that the technological parameters can be stabilized within the optimal interval all the time when the production process is interfered by raw material fluctuation, load change and the like.

The invention discloses an oil field injection-production collaborative optimization method and system fusing physical information reinforcement learning, and belongs to the crossing field of petroleum engineering and intelligent control. The method comprises the following steps: firstly, collecting oil field historical production data and carrying out normalization preprocessing; secondly, constructing a bidirectional long-short-term memory network based on an attention mechanism as an oil reservoir digital twin environment model for predicting a future production state; meanwhile, a physical information neural network based on a reservoir capillary force-saturation physical limit curve is innovatively introduced to serve as a physical constraint model, and the nonlinear constitutive relation between the water injection driving pressure and the theoretical water holding capacity of the micro-pores is analyzed; further constructing a composite reward function containing physical consistency penalty, and explicitly embedding the physical security boundary into a strategy optimization process of a depth deterministic strategy gradient algorithm; and finally, through interaction of the intelligent agent, the digital twin environment and the physical constraint model, a water injection strategy considering yield maximization and geological safety is output. According to the method, a data driving algorithm and an oil reservoir seepage mechanism are fused, so that the risk that a pure AI model easily generates physical illusion under sparse data and causes non-Darcy flowing water channeling is effectively avoided, and intelligent and safe collaborative optimization of an oil field injection and production system is realized.

The invention belongs to the technical field of ferrous metallurgy, and relates to an RH decarburization parameter prediction method based on data-driven algorithm and mechanism model coupling, and the method comprises the steps: collecting actual industrial parameters of an RH refining process; establishing a mechanism model; setting an initial value range of fitting parameters in the mechanism model; sampling in the initial value range to generate a plurality of groups of parameters, and inputting the parameters into the mechanism model for calculation to obtain an initial data set of carbon and oxygen mass fractions changing along with time; the initial data set is preprocessed; training a data driving algorithm by using the preprocessed data; optimizing hyper-parameters of the data driving algorithm to obtain an optimal data driving algorithm; and inputting the industrially measured mass fractions of carbon and oxygen into the optimal data driving algorithm, and outputting the optimal value of the predicted fitting parameter. The method has the advantages that two-way optimization of fitting parameter prediction and decarburization curve simulation is achieved through coupling of a mechanism model and a data driving algorithm, and prediction accuracy is remarkably improved.

The invention relates to the technical field of photovoltaic module fault diagnosis and management, and discloses a multi-modal data-driven photovoltaic module intelligent fault diagnosis method and system, and the method comprises the steps: collecting the electrical parameters, environment variables and physical characteristic data of a photovoltaic module in real time, carrying out the preprocessing, and outputting the processed standardized data; constructing a multi-type database, storing the processed standardized data, extracting a plurality of fault feature vectors in combination with a mixed fault detection model of a physical mechanism model and a data-driven algorithm, comparing the fault feature vectors with corresponding fault feature vector thresholds, and judging fault types; based on the fault feature vectors and the component parameters, fault inversion is carried out through a COMSOL multi-physics field simulation model, and a fault position is determined; and the fault types and the fault positions are collected, and a diagnosis result is output in a visual mode. Reliable technical support is provided for intelligent operation and maintenance of the photovoltaic power station, and the accuracy and timeliness of fault diagnosis are remarkably improved.

The invention relates to the technical field of driving training, and particularly discloses a driving training active braking system based on a data driving algorithm, and the system comprises a data collection module which is used for collecting the state data of a driving training vehicle, the operation behavior data of a driver, the vehicle surrounding environment perception data and other multi-source data in real time; the central processing and algorithm module is in communication connection with the data acquisition module and is used for fusing the vehicle state data, the driver operation data and the environment perception data and driving a risk assessment model to process based on the data so as to output a brake control instruction; the active braking execution module is in communication connection with the central processing and algorithm module, receives the braking control instruction and drives an execution mechanism connected with a vehicle braking system to actively brake the vehicle; and the man-machine interaction and data management module is bidirectionally connected with the central processing and algorithm module, provides information interaction for coaches and trainees, and stores whole-process data for post redisk and teaching analysis.

The application provides a kind of based on data-driven model predictive control's excavator bucket control method, belongs to data-driven predictive control technical field, comprising the following steps: S1, acquisition excavator system offline input-output data;S2, according to the data in S1, using least square method to identify the model parameters of excavator system, each group of model parameters is arranged in turn according to row, data set is constructed;S3, online update model parameters;S4, according to the parameters in S3, through MPC controller controls excavator bucket operation.The application combines data driving and model predictive control, effectively solves the problem that system model change affects the control effect of model predictive control, expands the application range of model predictive control;At the same time, the application improves the data-driven algorithm, introduces convex optimization theory in the data-driven algorithm, and optimizes the distance and weight of calculation. Since the optimal distance is calculated, more accurate model parameters can be obtained.

The invention discloses a vehicle body vibration fatigue acceleration test method based on multi-source features and an intelligent algorithm in the technical field of rail transit, and the method comprises the steps: collecting multi-source vibration load data in the service process of a vehicle body, and extracting the multi-dimensional features of the multi-source vibration load data; constructing a feature space, carrying out dimensionality reduction and clustering optimization on the feature space, and constructing a damage equivalent model by adopting a four-stage physical modeling process based on the optimized feature space; finite element analysis and a data driving algorithm are combined to realize accurate fitting of a damage mapping relation, an intelligent algorithm is adopted to carry out multi-objective optimization on a load working condition corresponding to an optimized feature space, and a test speed-up ratio and damage fidelity are balanced. An optimal acceleration test load spectrum is generated, a closed-loop test control system is built based on the acceleration test load spectrum, the test is executed in combination with the digital twin platform to complete a vehicle body vibration fatigue acceleration test and life evaluation, the test period is remarkably shortened, and the test efficiency is remarkably improved.

The invention provides a hydraulic characteristic prediction method and system based on physical knowledge and data dual drive, relates to the field of intelligent water conservancy, and solves the technical problem that an existing hydraulic characteristic prediction method is difficult to ensure physical consistency and prediction robustness at the same time. The method comprises the steps that multi-dimensional hydraulic data are collected and preprocessed, and a hydraulic database is constructed; constructing a physical driving model based on the hydraulic database, and optimizing the physical driving model by using a data driving algorithm; constructing a data prediction model based on the hydraulic database, and fusing the data prediction model and the physical driving model to construct a physical-data dual-driving fusion model; constructing a dual-constraint loss function of the physical-data dual-drive fusion model; and predicting hydraulic characteristic parameters based on a physical-data dual-drive fusion model. The method is used in the hydraulic characteristic prediction process.

The application discloses a financial data causality mining method and system for intelligent auditing. The method comprises the following steps: obtaining and preprocessing structured financial data and unstructured text data; fusing domain knowledge and data-driven algorithms to construct a financial causality graph; using a large language model to identify and quantify management intervention variables; constructing a structured causality model and performing counterfactual reasoning to obtain financial indicator prediction values under a non-intervention scenario; comparing actual observation values with prediction values, quantifying manipulation effects, and generating risk clues. The application also discloses a corresponding system, which comprises functional modules such as data acquisition and preprocessing, financial causality graph construction, management intervention variable identification and quantification, counterfactual reasoning, and manipulation effect mining. The application can realize analysis from abnormality discovery to causality attribution, quantifying the net influence of management decisions on financial results, thereby improving the identification ability and auditing efficiency of complex profit manipulation behaviors.