CN-121997730-A - Multi-physical field coupling-based camera-related network test simulation platform

CN121997730ACN 121997730 ACN121997730 ACN 121997730ACN-121997730-A

Abstract

The application relates to a simulation platform for a grid-related test of a camera based on multi-physical field coupling, which comprises the following components: the system comprises a power grid fault generating device, an electromagnetic field theory calculating device, a thermal field theory calculating device, a structural field calculating device and a multi-source input coupling device. The multi-physical field coupling simulation platform constructed by the scheme thoroughly overcomes the limitation of the traditional isolated calculation method through a real-time interaction mechanism of a dynamic integrated electromagnetic field-thermal field-structural field. Under the power grid fault working condition, the platform accurately simulates strong coupling effects among electromagnetic loss, temperature distribution and mechanical deformation (such as local overheating caused by eddy current loss, air gap offset caused by thermal expansion, and closed-loop influence of electromagnetic unbalance aggravated by eccentric quantity), so that the output actual parameter set (electromagnetic force density, temperature rise curve, air gap eccentric quantity and the like) is highly approximate to the real running state of the phase modulation unit.

Inventors

ZHAO HAIJIANG
ZHANG QIANG
LIU GUANSONG
ZHAO JUN
MA GUANGJUN

Assignees

中电投新疆能源化工集团托里有限责任公司

Dates

Publication Date: 20260508
Application Date: 20260116

Claims (10)

1. The grid-related test simulation platform of the phase regulator based on the multi-physical field coupling is characterized by comprising a grid fault generating device, a power grid test simulation device and a power grid test simulation device, wherein the grid fault generating device is used for providing a grid power supply environment of a phase regulator unit and simulating fault waveforms; The electromagnetic field theory calculating device is used for calculating the theoretical electromagnetic force density, the theoretical electromagnetic loss and the theoretical torque pulsation which are output by the phase modulation unit when the current of the fault waveform is input based on a transient finite element method; The thermal field theory calculating device is used for generating a theoretical temperature field and a theoretical temperature rise curve of the phase modulation unit according to the theoretical electromagnetic loss and the fluid heat exchange coefficient based on a transient heat conduction-convection equation; The structural field calculation device generates theoretical deformation characteristic data and theoretical air gap eccentricity of the phase modulation unit according to the theoretical electromagnetic force density and the theoretical temperature field; The multi-source input coupling device generates an actual parameter set according to the theoretical parameter set; The theoretical parameter group comprises theoretical electromagnetic force density, theoretical electromagnetic loss, theoretical torque pulsation, theoretical temperature field distribution, theoretical temperature rise curve, deformation characteristic data and air gap eccentric coupling, The actual parameter set comprises actual electromagnetic force density, actual electromagnetic loss, actual torque pulsation, actual temperature field distribution, actual temperature rise curve data, actual deformation characteristic data and actual air gap eccentric coupling; and the safety boundary calculating device is used for generating the fault probability of the phase modulation unit under the corresponding fault waveform based on the actual parameter set, calculating the fault probability of the phase modulation unit under different fault waveforms and generating a simulation experiment report.
2. The multi-physical field coupling-based camera grid-related test simulation platform of claim 1, wherein the electromagnetic field theory calculation means comprises: the coordinate generation module is used for determining an origin on the motor shaft of the phase modulation unit, taking the z axis as the axial direction of the rotating shaft, and enabling the planes of the x axis and the y axis to be perpendicular to the z axis so as to establish a three-dimensional rectangular coordinate system; the phase modulation unit response calculation module is used for calculating stator end current, rotor mechanical angle and exciting current of the phase modulation unit according to the fault waveform; And the electromagnetic field theory calculation module is used for calculating the theoretical electromagnetic force density, the theoretical electromagnetic loss and the theoretical torque pulsation according to the stator end current, the rotor mechanical angle and the exciting current.
3. The multi-physical field coupling-based camera grid-related test simulation platform of claim 1, wherein the thermal field theory calculation means comprises: The temperature information loading module is used for loading the heat exchange efficiency of each position of the heat exchange system; the temperature calculation module generates a theoretical temperature field and a theoretical temperature rise curve according to electromagnetic loss, initial temperature distribution and heat exchange efficiency of each position; The theoretical temperature field is the temperature distribution of each area in the three-dimensional rectangular coordinate system along with time, and the theoretical temperature rise curve is the temperature rise rate of key points in the three-dimensional rectangular coordinate system along with time.
4. The multi-physical field coupling based camera screening test simulation platform of claim 1, wherein the structural field computing means comprises: The structure information loading module is used for loading the initial geometric state of the phase modulation unit; and the structural information calculation module is used for generating theoretical deformation characteristic data and theoretical air gap eccentricity according to the initial geometric state, the theoretical temperature field and the electromagnetic force density.
5. The multi-physical field coupling-based camera screening test simulation platform of claim 1, further comprising a correction device; The correction device includes: The electromagnetic correction module is internally provided with a first correction neural network model and is used for correcting the calculated theoretical electromagnetic force density, theoretical electromagnetic loss and theoretical torque pulsation; The thermal field correction module is internally provided with a second correction neural network model and is used for correcting the calculated theoretical temperature field and the theoretical temperature rise curve; the structure correction module is internally provided with a third correction neural network model and is used for correcting the calculated theoretical deformation characteristic data and the theoretical air gap eccentricity; the first correction neural network model, the second correction neural network model and the third correction neural network model are used for correcting errors of theoretical calculated values and actual values to obtain correction parameters.
6. The multi-physical field coupling based camera screening test simulation platform of claim 5, wherein, The first modified neural network model is a time sequence feature extraction network based on bidirectional LSTM; the second modified neural network model is a multi-scale feature fusion network based on 3D convolutional coding; the third modified neural network model is a robust shallow network based on feature expansion.
7. The simulation platform of the network-related test of the camera based on the multi-physical field coupling, which is characterized in that training data of the first modified neural network model, the second modified neural network model and the third modified neural network model are theoretical parameters and actual parameters acquired when the phase modulation unit is in a steady state, wherein the theoretical parameters are used as characteristic data, and the actual parameters are used as labeling data; In the data for training the first correction neural network model, the change amplitude of the theoretical temperature field, the theoretical temperature rise curve, the theoretical deformation characteristic data and the theoretical air gap eccentricity is maintained within a preset range; In the data of training the second correction neural network model, the variation amplitude of electromagnetic force density, theoretical electromagnetic loss, theoretical torque pulsation, theoretical deformation characteristic data and theoretical air gap eccentricity is maintained in a preset range; in the data for training the third modified neural network model, the electromagnetic force density, the theoretical electromagnetic loss, the theoretical torque pulsation, the electromagnetic force density, the theoretical electromagnetic loss and the variation amplitude of the theoretical torque pulsation are maintained in a preset range.
8. The multi-physical field coupling-based camera screening test simulation platform of claim 1, wherein the multi-source input coupling device comprises: The information input module is used for acquiring a theoretical parameter set X; wherein, the method comprises the steps of, The representation of the empty set is made, 、、 Respectively representing theoretical electromagnetic force density, theoretical electromagnetic loss and theoretical torque pulsation; 、 respectively representing a theoretical temperature field and a theoretical temperature rise curve; 、 Respectively representing corrected theoretical deformation characteristic data and theoretical air gap eccentricity; the data enhancement module is used for labeling the real contribution of each parameter in the theoretical parameter set X to each parameter in the actual parameter set based on the Group Lasso model to obtain contribution information; and the large model coupling calculation module is used for inputting the theoretical parameter set X and the contribution information, and searching the coupling relation between the characteristic parameters based on the multi-channel cross network so as to generate an actual parameter set.
9. The simulation platform for the grid-related test of the camera based on the multi-physical field coupling as claimed in claim 8, wherein the obtaining mode of the contribution information comprises the following steps: s1, taking all collected theoretical parameter groups X as an observation variable matrix XT and taking all collected actual parameter groups as a response variable moment YT; The dimension of the observation variable matrix XT is n×p, the dimension of the response variable matrix YT is n×p, n represents the number of collected samples, and p=7; S2, establishing an influence group for each element in the actual parameter set Y Where b denotes the index of the element in the actual parameter set Y, b.epsilon.1, 7, l is the index affecting the packet, B and l are positive integers, A total number of influencing packets representing the b-th parameter of the actual parameter set; s3, obtaining all influence groups Setting each influence packet Is used for initializing coefficient vectors of (a) G represents index of influencing grouping, k represents iteration number, setting initial step size Step size reduction factor Coefficient of descent Loss function ; S4, carrying out the following iterative operation on each influence group until convergence, wherein when each influence group carries out the iterative operation, the coefficient vectors of the rest influence groups are fixed; according to the current coefficient vector Gradient calculation for group design matrix X g And a hessian scalar h g ; ; ; Representing a preset loss function, g representing the index of the influencing group, and X g = the column index in the observation variable matrix XT belonging to the influencing group Is used for the matrix of the matrix, Representing the gradient of the loss function, Representing the sub-vectors corresponding to the first group in the complete coefficient vectors of the kth iteration; ; representing the loss function with respect to the g-th group of coefficient vectors, The minimum curvature constant is indicated as such, Representing the approximate scalar of the hessian matrix block, max representing the vector maximum operation, diag representing the matrix diagonal extraction operation; calculating an auxiliary vector u; ; s5, judging the updating direction of the coefficient vector according to the auxiliary vector; ; Indicating the direction of the update, Represents a penalty threshold, λ represents a weight coefficient; s5, calculating the reference quantity according to the updating direction Calculating the optimal step length according to the reference quantity : ; Indicating the amount of the desired drop in the liquid, Representing matrix transposed symbols; S6, according to the optimal step length Updating the coefficient vector; ; s6, judging whether iteration is stopped according to a convergence condition, wherein the convergence condition is as follows: Stopping if the iteration condition is met, and continuing iteration if the iteration condition is not met; s7, obtaining coefficient vectors of all influence groups The coefficient vector is used as the contribution information.
10. The multi-physical field coupling based camera grid-related test simulation platform of claim 9, wherein the multi-channel crossover network comprises: An input layer for inputting a theoretical parameter set X and contribution information W; The feature embedding layer projects the input theoretical parameter set X to a high-dimensional space through the full-connection layer to generate a high-dimensional vector: The cross network is CrossNet, the weight matrix of the cross network is initialized by the contribution information W, a fine tuning boundary is set, and a high-dimensional vector is input into the cross network to generate and display high-order characteristics; The full-connection network is a three-layer full-connection network, and implicit characteristics among input theoretical parameter sets X are extracted; the feature fusion layer fuses and displays the high-order features and the hidden features based on a gating fusion mechanism and outputs the fused features; And the multitasking output layer generates an actual parameter set according to the output fusion characteristics.

Description

Multi-physical field coupling-based camera-related network test simulation platform Technical Field The application relates to the technical field of high-voltage transmission, camera simulation and physical field coupling, in particular to a camera grid-related test simulation platform based on multi-physical field coupling. Background The statements in this section merely provide background information related to the present disclosure and may not constitute prior art. The phase modulation unit is used as a key reactive compensation and voltage support device in a modern power system, and the operation performance of the phase modulation unit is directly related to the stability, the safety and the electric energy quality of a power grid. It is important to perform accurate simulation tests before the equipment is put into actual operation. The simulation test has the core effects of simulating the running state of the power grid under various working conditions (such as load fluctuation, fault disturbance, system oscillation and the like) and evaluating the dynamic response characteristic, reactive power regulation capability, overload margin and stress level of key components of the phase modulation unit. Through simulation, the behavior of the equipment in the actual power grid environment can be predicted, the effectiveness of design parameters and control strategies of the equipment is verified, and potential risks are identified. At present, aiming at simulation test of a phase modulation unit, a commonly adopted method is to calculate the adjustment limit of each physical property of the phase modulation unit in isolation. This mainly includes: and (3) calculating the highest allowable temperature rise of key components such as stator windings, rotor windings, iron cores and the like under specific cooling conditions based on a thermodynamic model. And the electromagnetic limit is that the maximum output capacity (such as maximum capacity/inductive reactive power) under the constraints of stator and rotor magnetic circuit saturation, exciting current upper limit, end magnetic leakage and the like is calculated according to an electromagnetic field theory. And (3) deformation/mechanical limit, namely calculating the maximum allowable deformation or stress level of a rotor deflection, bearing vibration and key structural components (such as an end cover and a base) under the comprehensive actions of electromagnetic force, thermal stress and centrifugal force by taking the mechanical properties of materials into consideration. These limit values are usually calculated separately under idealized or simplified boundary conditions. When the phase modulation unit operates in an actual power grid, physical fields such as a temperature field, an electromagnetic field, a stress field (deformation/vibration) and the like are highly coupled and mutually influenced. For example, excessive excitation current (electromagnetic limit) can lead to a sharp rise in winding temperature (temperature limit), which in turn can exacerbate material expansion and mechanical deformation (deformation limit), thereby affecting air gap uniformity, which in turn worsens electromagnetic performance and vibration levels, creating a complex dynamic feedback process. The existing method for calculating each single limit in an isolated way cannot fully consider the multi-physical field intensity coupling effect, so that the calculation result is often over-ideal. Therefore, the "working limit" determined based on these single limits cannot truly reflect the comprehensive performance boundary of the phase-modulation unit under the actual power grid conditions of complex, dynamic and multi-constraint coupling. The limitation makes it difficult to accurately predict the actual bearing capacity and safety margin of the equipment in the extreme or transient process of the power grid, which may cause misjudgment of the running risk of the equipment or underestimation of the potential of the equipment, and cannot provide accurate and reliable simulation basis for safe and stable running of the power grid. Disclosure of Invention The summary of the application is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. The summary of the application is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Some embodiments of the application provide a multi-physical field coupling-based camera grid-related test simulation platform to solve the technical problems mentioned in the background section. As a first aspect of the present application, some embodiments of the present application provide a multi-physical field coupling-based grid-related test simulation platform for a camera, including: the power grid fault generating device is used for providing a po