68 results about "Model order" patented technology

The invention discloses a high arch dam operation modal parameter automatic identification method and system based on discharge excitation. The method comprises the following steps: 1) obtaining a vibration displacement response signal: determining a penalty factor and an optimal decomposition layer number based on an adaptive multivariate variational mode decomposition algorithm to obtain an optimal IMF component of each sensor channel signal, and performing IMF component screening and signal reconstruction through a frequency domain cross correlation coefficient to realize adaptive noise reduction of the signal; 2) establishing a Monte Carlo three-dimensional stability diagram based on a random subspace recognition algorithm driven by a covariance matrix; and 3) modal parameter automatic identification based on an intelligent clustering algorithm. According to the method, the noise is suppressed by automatically optimizing the modal component reconstruction signal of the multi-sensor vibration signal; a Monde-Carlo three-dimensional stability diagram is established in combination with a Monde-Carlo theory and a covariance driven random subspace method to determine a model order, and automatic interpretation of the stability diagram is realized by applying improved fuzzy clustering, so that operation modal parameters of the high arch dam are accurately identified.

The invention discloses a mining area load spectrum anomaly detection method based on standard variable analysis, and belongs to the technical field of hydraulic pump anomaly detection.The mining area load spectrum anomaly detection method comprises the steps that S1, standardized data are obtained through data preprocessing, and a typical working condition data set of a hydraulic pump is constructed in combination with working conditions of an excavator; s2, constructing a historical vector and a future vector, and constructing a historical observation matrix and a future observation matrix according to the historical vector and the future vector; s3, constructing a Hankel matrix according to the autocorrelation matrix and the cross-correlation matrix, decomposing the Hankel matrix and determining a model order; s4, mapping the original data to a standard variable space and a residual space, and respectively evaluating the total variable quantity of the standard variable in the state space and the sum of squares of change errors in the residual space; and S5, determining an evaluation threshold value, and if the evaluation threshold value exceeds the control line, judging that the hydraulic pump operates abnormally. The method is based on standard variable analysis, adopts the pressure pulsation data of the hydraulic pump to perform anomaly detection, is sensitive to the internal running state of the pump, is not easily influenced by the external environment, and can perform early warning on faults of the hydraulic pump.

The invention relates to a vehicle adaptive model prediction control method. The method comprises the following steps: collecting vehicle dynamic parameters, and constructing a two-degree-of-freedom vehicle dynamic model according to the dynamic parameters; constructing a first matrix used for describing an input and output relation of the linear state space equation; determining a model order of a vehicle model corresponding to the two-degree-of-freedom vehicle dynamics model through an akaike information criterion, and identifying parameters of the first matrix through a recursive least square method of introducing a forgetting factor to obtain the vehicle model; self-adaptively adjusting forgetting factors in the vehicle model through fuzzy logic and identification residual errors to obtain an optimized vehicle model; and establishing a vehicle controller based on the optimized vehicle model to control a vehicle control system in real time. According to the method, effective tracking of the expected state of the vehicle under different working conditions can be realized, the identification speed and the modeling precision are improved, and the modeling complexity is reduced.

The invention discloses a generator outlet circuit breaker temperature rise simulation method under a model order self-adaptive strategy. A three-dimensional model for finite element simulation of a generator outlet circuit breaker is established; performing temperature field finite element simulation based on the three-dimensional model to obtain temperature data of the generator outlet circuit breaker under different working conditions; sampling the temperature data to establish a sample point data matrix; performing order reduction processing on the sample point data matrix to generate an order-reduced data matrix; using the reduced-order data matrix to train the proxy model to generate a reduced-order model; when an error between a temperature simulation value output by using the reduced-order model and a temperature measured value meets a constraint, simulating temperature rise of the generator outlet circuit breaker by using the reduced-order model, and if the error does not meet the constraint, performing order expansion on the reduced-order model; and when the error meets the constraint, simulating the temperature rise of the generator outlet circuit breaker by using the order-expanded reduced-order model. The dimension of the reduced-order model is automatically determined and dynamically adjusted, and the real-time performance and the accuracy of the GCB temperature rise simulation reduced-order model are effectively considered.

A method for identifying a model for modeling cable stress relaxation dynamics of an imprint head is provided. The method includes performing a non-contact imprint test run for a predetermined number of wafers. Data from a force trace and a position trace of the imprint head during the non-contact imprint test run are collected to compute cable stress relaxation forces. A model having a plurality of different model orders is generated. A set of parameters of the model for each model order is identified. Residual sum of squares for wafer average error for each of the model orders is calculated based on the obtained cable stress relaxation forces. The set of parameters of one of the plurality of model orders may be based on a difference in the residual sum of squares for wafer average error between the one of the plurality of model orders and a next higher model.

The embodiment of the invention relates to a model sample generation method and device, equipment and a storage medium. The method comprises the steps that based on input keywords, a plurality of question and answer pairs in a natural language are sequentially generated through a first machine learning model, and a second question and answer pair in the multiple question and answer pairs is generated at least based on replies in a first question and answer pair generated before the second question and answer pair; integrating the plurality of question and answer pairs by using a second machine learning model, and generating a multi-hop question aiming at the question and answer model and a reference answer aiming at the multi-hop question; and in response to determining that the multi-hop question and the reference reply meet the first quality requirement, generating a test sample and / or a training sample of the question and answer model.

The application discloses a kind of manifold curved surface model ordered edge chain fast solving method, object is the three-dimensional model represented based on the manifold curved surface of triangular facet;Main content is: innovative edge chain solving train of thought, by prior judging the maximum endpoint of manifold curved surface model in ±X, ±Y, ±Z six directions is combined with odd-even respectively traversed mode, first edge point is quickly solved, effectively improve solving speed;Precise edge point criterion is proposed, in combination with the adjacent relationship of first edge point and other edge points, in turn recursively, directly determine edge chain, avoid the time and power waste problem of complete traversal, improve solving speed while the edge chain solved from the sequence, without additional arrangement order;Proposed edge chain acquisition method under non-half edge data structure, based on multiple array instead of half edge data structure, completely solve the information redundancy and power waste problem caused by half edge data structure, widen the scope of solvable model, also provide prerequisite for algorithm cross language deployment.

The application discloses a kind of based on system disturbance response's self anti-interference controller setting method, comprising the following steps: determining the model order of controller and controlled object;According to model order, data vector and model parameter vector are set;Select the initial value of calculation of feedback gain matrix and data covariance matrix;Recursive calculation feedback gain matrix, data covariance matrix and the estimated value of the model parameter vector, until the estimated value of model parameter vector tends to be stable;According to stable value, calculate the coefficient estimate of controlled object model;Based on each coefficient estimate, design self anti-interference controller parameter.The method of the application can guarantee system closed loop and system set value unchanged, only according to the disturbance response of system setting self anti-interference controller parameter, overcome the problems existing in prior art, the method of the application has good tracking and anti-interference performance, help to further promote the field application of self anti-interference controller in thermal process control.

The application discloses a mixed order regional ionospheric model construction method, which comprises the following steps: establishing a mixed order regional ionospheric model based on observation data; according to whether the ionosphere is in a quiet period or an active period, using different orders and model parameter update intervals to establish corresponding regional ionospheric models; designing a model order self-adaptive adjustment strategy to realize switching of the mixed order regional ionospheric model to a first-order polynomial model or a second-order polynomial model; and unifying generation approaches of mixed order regional ionospheric model parameters to realize estimation of first-order / second-order polynomial model parameters. The mixed order regional ionospheric model has the advantages that, compared with the first-order polynomial method, the problem of low modeling accuracy in the ionosphere active period is overcome; compared with the second-order polynomial method, the limitations of model overfitting and modeling instability in the ionosphere quiet period are avoided; and in terms of parameter generation and modeling accuracy, the mixed order regional ionospheric model combines the advantages of the second-order model and the first-order model, thereby ensuring the accuracy and applicability of regional ionospheric modeling.

This invention provides a radar motion parameter estimation method and apparatus based on information criterion values. It acquires raw radar echo data, performs range processing and target extraction to obtain a target slow-time complex sequence. Based on the target slow-time complex sequence, it constructs multi-order motion candidate models, calculates the residual sum of squares, and calculates the information criterion value based on the residual sum of squares. The optimal model order is determined based on the information criterion value, and the target radial motion parameters are inverted and output based on the phase coefficients corresponding to the optimal model order. Instead of pre-fixing the use of a first-order, second-order, or third-order model, it determines the optimal model order after constructing multi-order candidate models and evaluating them using the calculated information criterion value. This approach adaptively determines the appropriate motion model order for the target based on the phase evolution characteristics of the signal itself, and completes motion parameter estimation accordingly. This ensures estimation accuracy while suppressing unnecessary model complexity, thus improving its engineering application value.

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.

The invention relates to a pipeline gas pressure and flow measurement method and system based on big data, and the method comprises the steps: a data collection module records the temperature, signal-to-noise ratio, signal gain and signal intensity of a flow meter sensor at different temperatures and different flows, builds a database to form a training sample set, defines a second-order polynomial regression model, and obtains a first-order polynomial regression model and a second-order polynomial regression model; s3, the training sample set is substituted into the model for fitting, a to-be-estimated regression coefficient is solved, the regression coefficient obtained in the step S3 is applied to the pipeline gas pressure model, data different from training data is substituted into the fitting equation set, a fitting pressure value is obtained, and when the error is larger than 1.5%, the model order is adjusted again, and fitting is conducted again; according to the invention, the gas pressure and flow of the pipeline are measured, the pipeline is prevented from being damaged, and the gas pressure and flow can be accurately measured.

The invention discloses a mode identification method based on an unsupervised optimization covariance random subspace method. The method comprises the following steps: arranging a vibration sensor on a to-be-detected structure with noise interference or a weak excitation mode to obtain a structure dynamic response; constructing a Hankel matrix, calculating a covariance matrix Ri, and constructing a Toeplitz matrix based on the Ri; calculating an extended observable matrix Oi and an extended controllable matrix Gamma i based on the weighted Toeplitz matrix; defining the range of the row block number i and the model order N of the Toeplitz matrix; analyzing the sensitivity of the row block number i of the Toeplitz matrix and the model order N based on a parameter optimization index kP (i, N); the singular entropy increment of the weighted Toeplitz matrix T1i is calculated; calculating the singular entropy increment curvature of the weighted Toeplitz matrix T1i, and determining a critical model order Nc (i); calculating an accumulated parameter optimization index kappa P value from a minimum model order Nmin to a critical model order Nc (i) under the condition of different Toeplitz matrix row block numbers i; selecting a parameter combination {iopt, Nopt} corresponding to the minimum parameter as an optimal parameter; identifying a system matrix and determining modal parameters, and drawing an original stability diagram; and on the basis of DBSCAN clustering, automatically identifying each order of physical modality from candidate modalities containing noise interference. A corresponding system is also disclosed.

The invention discloses a broadband signal measurement method based on improved Prony, and the method comprises the following steps: 1, carrying out the singular value decomposition of a broadband signal, determining an initial singular value demarcation point through an energy accumulation method, and introducing a Pearson correlation coefficient to distinguish the main frequency components and noise of the signal, thereby determining the model order of Prony; 2, preprocessing the broadband signal by using variational mode decomposition, solving a variational problem through iteration, decomposing the complex broadband signal into a plurality of intrinsic mode functions with narrowband frequency spectrums, and reducing the complexity of signal processing; step 3, introducing a disturbance matrix satisfying a Toeplitz structure, iteratively optimizing a linear prediction coefficient of a Prony algorithm, using the Prony algorithm after parameter optimization to identify each mode, obtaining frequency, amplitude and phase information of each mode, and realizing measurement of broadband signals; according to the method, the problem of insufficient measurement precision caused by noise interference and difficulty in model order selection in broadband signal measurement of a traditional Prony algorithm is solved.

A method for the model-based, high-resolution separation of radar targets for a radar sensor, in which the radar sensor initially generates radar data by capturing radar targets by sampling a field of view of the radar sensor. Various model orders are calculated for the number of radar targets with the aid of a grid-based method and are interleaved in one another. A highest model order is specified, and only the highest model order is calculated, so that the lower model calculations are implicitly produced from the calculation of the highest model order.

The invention provides a small disturbance stability estimation method for a grid-following type and grid-constructing type hybrid grid-connected converter, and relates to the technical field of power grids, and the method comprises the steps: aiming at each preset time period, based on the corresponding control and structure parameters of the grid-following type grid-connected converter and the grid-constructing type grid-connected converter in a small disturbance scene, calculating the stability of the grid-following type and grid-constructing type hybrid grid-connected converter; determining a system stable equilibrium point, a system unstable equilibrium point and equilibrium point stability conditions corresponding to the preset time period; determining a period stability analysis result corresponding to the preset time period based on a balance point linear differential equation, the system stable balance point, the system unstable balance point and the balance point stability condition; and determining a target stability analysis result based on the period stability analysis results corresponding to all the preset time periods. The method is applied to the condition of series-parallel connection of the following grid-connected converter and the construction grid-connected converter, is not influenced by a relatively high model order of the grid-connected converter, and can obtain an accurate stability analysis result.

The present invention provides a method for analyzing the stability of heterogeneous systems based on a simplified nonlinear dynamics model, comprising the following steps: establishing a simplified model topology based on the heterogeneous system and obtaining a phase angle difference, where the phase angle difference is the difference between the phase angle of a first node and the phase angle of a second node; obtaining a simplified nonlinear dynamics model based on the heterogeneous system and the phase angle difference based on the simplified model topology; solving the nonlinear dynamics model using an averaging method to obtain an analytical solution, where the analytical solution is an expression of the square of the system amplitude; differentiating the analytical solution with respect to time to obtain a third first-order derivative; obtaining a stability criterion when the third first-order derivative is zero; and determining the stability of the heterogeneous system based on the stability criterion. The method for analyzing the stability of heterogeneous systems based on a simplified nonlinear dynamics model provided by the present invention is simple and efficient, has a low model order, retains the nonlinear characteristics of the system, performs an analytical solution based on a generalized averaging method, proposes a system stability criterion, and facilitates stability analysis.

The invention discloses a model sequence quantification method and system based on channel balance and related equipment, and belongs to the technical field of model compression. According to the method, for a target operator in a quantitative model, output feature maps of a floating point model and the quantitative model under the same input are respectively obtained; respectively carrying out independent normalization processing on each channel of the two feature maps, calculating the similarity between the corresponding channels after normalization, and carrying out aggregation to obtain a channel balance evaluation index representing the quantitative fidelity; if the index does not reach the preset threshold value, the quantization precision of the layer is improved, the model is updated, and then the model is updated to re-execute the evaluation process; the process is iterated layer by layer according to a network topology sequence until all layers are processed. According to the method, through finer channel-level evaluation and dynamic error suppression on the quantization effect, the quantization bit width can be allocated more accurately, and the hardware resource utilization efficiency is improved while the model quantization precision is ensured.

The invention discloses a comprehensive modeling and multi-target complementary control method for a laser communication coarse tracking system, and belongs to the field of laser communication technology control, the laser communication coarse tracking system is used for control, and the method comprises the following steps: carrying out linear modeling on the coarse tracking system, combining a low-frequency band of a sine sweep frequency signal and a middle-high frequency band of a pseudo-random signal, obtaining a system frequency response, and obtaining a system transfer function through a Hankel matrix identification method; and performing nonlinear modeling on the coarse tracking system, establishing a Stribeck friction model, designing a feed-forward compensation algorithm based on the Stribeck friction model, and obtaining a multi-target complementary control method to control the coarse tracking system. According to the method, accurate identification of model orders and parameters is realized in the linear part, the low-speed tracking performance is remarkably improved in the nonlinear part, the contradiction between the system performance and robustness is effectively solved, and the tracking precision and the anti-interference capability are remarkably improved.

The invention discloses a generalized ARMA model-based gas turbine compressor blade tip clearance data modeling and prediction method, and relates to data modeling and prediction. The method comprises the following steps: collecting multi-working-condition operation data of a gas turbine, and screening characteristic parameters related to blade tip gap height by using a grey relational analysis method; constructing a multi-input generalized ARMA model, replacing the random interference variable of the traditional ARMA model with a screening variable, and introducing a ridge regression algorithm to solve parameters; determining an optimal model order and parameters by using a particle swarm optimization algorithm; and evaluating a model prediction effect by using an average absolute percentage error, and performing adjustment and optimization. The complex coupling relation between the blade tip clearance and the multi-working-condition parameters is effectively captured, and high-precision, efficient and economical technical support is provided for active control over the blade tip clearance of the gas turbine.

The application discloses a kind of long-acting time-varying monitoring methods for dam reinforcement based on Bayes real-time updating, comprising: S1. ARMA model is constructed to simulate dam state and determine model order;S2. using prior distribution and likelihood function, using Bayes method to calculate and obtain the posterior density function of each parameter of dam in running state;S3. using last posterior distribution as prior distribution in real-time calculation, repeat the S2 step, in the calculation process, least square method is used to realize model parameter estimation, variable forgetting factor is introduced, and recursive algorithm is used, different forgetting factor is used to past monitoring data, realize parameter adaptive adjustment.The application can adapt to the time-varying characteristics of dam working behavior, real-time fusion historical data and new monitoring information, more accurately and reasonably reflect the nonlinear time-varying relationship between influence quantity and effect quantity, so as to fully embody the typical time-varying characteristics of dam working behavior.

The invention discloses a motion prediction method for a ship-borne landing platform, and relates to the technical field of unmanned aerial vehicle landing, and the method comprises the following steps: S1, obtaining historical data, S2, estimating initial parameters, S3, analyzing the order of a model, S4, collecting real-time data, S5, updating model parameters, and S6, outputting prediction data. According to the method, the model parameter estimation values corresponding to different moments are calculated according to the sample data by using the recursive least square method, so that the finally determined model parameters can be more accurate, it is ensured that motion prediction during landing of the unmanned aerial vehicle can fit reality, the estimation error value corresponding to the current moment is analyzed by using the prediction analysis algorithm, and the prediction accuracy of the unmanned aerial vehicle is improved. The optimal model order is calculated according to the estimated error value, the model parameters and the order value are transmitted to the time sequence model, the model is used for accurately estimating the motion amount of the unmanned aerial vehicle on the ship-borne landing platform at the next moment, and the calculated amount required by motion prediction is simplified to a certain extent.

This invention discloses a method, system, and device for identifying subsynchronous oscillation modes under environmental excitation, belonging to the field of power system oscillation detection. Kalman filtering (KF) is used to estimate the coefficients of an autoregressive (AR) model to fit measured environmental data. The damping factor and frequency of the subsynchronous oscillation (SSO) mode can be directly obtained by solving for the roots of the characteristic polynomial corresponding to the AR model. Furthermore, multi-model segmentation filtering (MMPF) is applied to KF to select the correct AR model order. This invention combines the KF algorithm with MMPF, enabling the algorithm to directly identify weak SSO mode parameters from environmental data. The method has simple operation steps and short response time, thus possessing good application prospects.

An adaptive engine and adaptive control method. The method comprises receiving an input regressor and applying one or more estimation laws to the input regressor to estimate a plurality of estimated model parameter tensors for a nonlinear model. The method also includes computing an estimation error or a cost function using a system output measurement and an estimated system output for the previous iteration and determining a model order based at least in part on the estimation error or the cost function and receiving two or more possible control signals each using a corresponding control portion of the nonlinear model. In addition, the method includes generating two or more estimated system outputs each using a corresponding estimation portion of the nonlinear model and selecting a preferred control signal from a set comprising at least the two or more possible control signals or a preferred combination of possible control signals.

The invention discloses a seasonal rural power distribution network load prediction method and system based on kernel density estimation. A power distribution network load sample is collected; carrying out stable processing on the load sample sequence by adopting a run test method to obtain a stable load sample sequence; based on a kernel density estimation method, establishing a load random prediction model of the seasonal rural power distribution network by using the stable load sample sequence; establishing a cross validation function to meet the minimum model order of the cross validation function, and taking the minimum model order as the optimal lag order; calculating a confidence interval of the short-term load prediction value; and obtaining a short-term load prediction value based on the load random prediction model under the optimal lag order, and taking the short-term load prediction value in the confidence interval as a seasonal rural power distribution network load prediction result. The problem of load instability is solved, more accurate load prediction is realized, and the reliability of power supply is improved.

The application provides a digital service market simulation method and system, which models an external demand environment in a digital service market, maps the external demand environment to order states of different types and different difficulties representing user demand, and sets an order arrival strategy and an order difficulty generation strategy; models a subject of a digital service system in the digital service market, the subject including a platform, an organization and an execution agent; models order delivery rules between organizations in the digital service market and execution agent cooperation rules within the organizations; and configures external demand environment, subject and rule parameters of a target digital service market based on modeling results of the three parts, and performs simulation. The application realizes mapping of the digital service market to a computer simulation system through modeling from three aspects of an external environment, a subject and rules, and solves the problem of lack of intuitive and effective simulation methods and experimental systems in the existing digital service market.

The invention discloses a method for predicting the rotating speed of a power generation steam turbine based on differential thermal inertia, and aims to overcome the defects of thermal inertia processing homogenization, easy drifting of dimensionless numerical values, low feature screening precision, poor model adaptability and the like of an existing method for predicting the rotating speed of a steam turbine. A prediction system is constructed through four core innovations: a differential thermal inertia lag step number adaptation formula is proposed, lag time is set in a grading manner based on thermal parameter physical characteristics, and a sampling interval is adapted; a full-process dimensionless number [0-1] forced constraint formula is designed, and the numerical drift problem is thoroughly solved; establishing a three-dimensional IMF component screening formula, and fusing correlation, error and frequency characteristics to purify effective characteristics; and constructing a residual-driven self-adaptive ARIMAX model formula to realize dynamic adjustment of the order of the model. The method is suitable for various sampling rate scenes, the prediction precision and stability are obviously better than those of the prior art, and the method can be widely applied to power generation turbine rotating speed real-time prediction and unit regulation and control and provides support for safe operation of a unit.

The invention belongs to the technical field of power system simulation and information interaction, and particularly relates to an information interaction method for joint simulation of a main network and an active distribution network, and the method comprises the steps: building a multi-rate clock synchronization and prediction compensation mechanism, and achieving the coordination of the simulation step size of the main network and the distribution network and the perspective compensation of boundary data; constructing a hierarchical boundary information model containing static topology, dynamic electrical quantity and aggregation characteristic data, and supporting high-fidelity equivalent representation of dynamic external characteristics of the distributed power supply; executing a self-adaptive interaction strategy based on the running state, and dynamically adjusting the interaction frequency, the data granularity and the model order according to the normal, transient or topological change state; boundary data consistency verification and closed-loop correction are implemented, and numerical value stability is ensured through sensitivity information exchange and iterative correction. According to the invention, the precision, efficiency and convergence of joint simulation are improved.

The invention discloses a power system inertia on-line evaluation method and device. The method comprises the steps that a measurement data signal of a grid-connected device is collected and preprocessed; determining the occurrence moment of large disturbance of the system from the preprocessed data, and extracting observation data; establishing a controlled autoregression model according to the observation data, and determining a model order; a self-adaptive two-factor robust recursive least square algorithm is adopted, a robust factor is introduced into the least square algorithm, model parameters are recognized and identified, and a transfer function of each device is obtained; impulse response is applied to the transfer function of each device, a response initial value is calculated to determine the inertia of the device, and the total inertia of the system is obtained after aggregation. According to the invention, the system disturbance is monitored on line through the DFA, the observation data range for inertia evaluation is accurately judged, and the real-time performance and accuracy of inertia evaluation are ensured.

The invention discloses a fault diagnosis method and device for an overhead line, terminal equipment and a storage medium, and belongs to the technical field of power transmission lines, and the method comprises the steps: carrying out the frequency domain analysis of a frequency response function, generating a modal frequency and a damping ratio pole under each model order, constructing a stability diagram, and carrying out the fault diagnosis of the overhead line according to the stability diagram. Determining a frequency pole set of each model order in a preset frequency interval and a frequency variation coefficient; according to the frequency variation coefficient, determining a frequency tolerance threshold value of each model order, and according to the damping ratio pole, the damping ratio tolerance threshold value and the frequency tolerance threshold value, identifying a stable pole and a stable cluster of each model order in the stability map; calculating a current modal parameter based on each stable cluster; when it is determined that the overhead line has the fault risk, an alarm signal is sent out, and therefore the defect that the fault diagnosis accuracy of the overhead line is low due to the fact that modal parameter identification cannot be accurately conducted through an existing method is overcome.