64results about How to "Quick forecast" patented technology

The invention relates to the technical field of composite material performance prediction, and discloses a composite material flame retardant property and thermal stability prediction method based on thermogravimetric analysis and infrared spectrum data fusion. The invention aims to solve the problem that it is difficult to effectively fuse multi-modal data to realize quantitative prediction in the prior art. The method comprises the following steps: firstly, collecting thermogravimetric and infrared spectrum combined data, calculating gas transmission lag time, and carrying out forward correction on infrared spectrum data to realize time sequence alignment; then solid-phase thermogravimetric feature vectors and gas-phase infrared feature vectors are extracted, and a fusion feature matrix is generated; extracting solid-phase and gas-phase characteristics respectively by using a dual-channel LSTM network, and performing weighted fusion based on a self-attention mechanism; and finally, synchronously outputting a flame retardant index predicted value and a flame retardant grade classification result through numerical mapping and classification judgment. According to the method, multi-dimensional feature mining of the whole pyrolysis process of the material is realized, and a rapid and accurate digital evaluation means is provided for screening of the flame-retardant material.

The present application relates to a kind of lake reservoir type chlorophyll concentration prediction method, device and medium, method includes the following steps: obtaining the water quality, water dynamics and weather of the online automatic monitoring data of lake reservoir including forecast point and upstream point;The online automatic monitoring data is preprocessed, and the data after processing is obtained;The data after processing is carried out feature extraction to obtain the water quality feature and water dynamics index feature of forecast point and upstream point, and construct cumulative illumination feature;Upstream water quality feature, water dynamics index feature and illumination feature are input into the fusion prediction model pre-trained, to obtain upstream transport chlorophyll prediction result and time series influence chlorophyll prediction result, using error reciprocal method upstream transport chlorophyll prediction result and time series influence chlorophyll prediction result are fused;The chlorophyll prediction result obtained by fusion is output.Compared with prior art, the present application has the advantages of high accuracy, strong stability and the like.

The application discloses a method for predicting the thrust of a single-side expansion nozzle in a test state. First, at least three groups of performance data of the single-side expansion nozzle are obtained through simulation or test, and the data includes total pressure, total temperature, back pressure, flow rate and thrust coefficient of the nozzle. Second, the total pressure and the total temperature in each group of simulation or test are ensured to be the same, and if the total pressure and the total temperature are not the same, the total pressure and the total temperature are converted into the same total pressure and total temperature by using a formula. Then, a y=Ae ‑Bx +C function is introduced to establish a continuous relationship between the converted nozzle thrust and the back pressure, and the data at any point in the sample range is known. Only the relationship between the thrust coefficient and the pressure ratio is considered, and the influence of the specific heat ratio is ignored, so that the thrust coefficient is considered to be equal as long as the pressure ratio of the nozzle is the same. The nozzle thrust in the test state may be affected by the high-altitude cabin structure, and the application provides correction relationship formulas of different factors, so that the full-flow field thrust prediction of the nozzle can be more accurately constructed.

This invention relates to the field of navigation and timing technology, specifically to a GNSS timing data prediction method and related equipment. The method involves acquiring Roland timing data when GNSS timing data is invalid and preprocessing it. The preprocessed data is then input into a trained single-hidden-layer extreme learning machine model. This model calculates the hidden-layer output matrix by randomly initializing the input layer weight matrix and bias vector, and solves for the output weights using regularized least squares. The model outputs a normalized GNSS timing data prediction value, which is then denormalized to obtain the final GNSS timing prediction result. This method combines the high stability of the Roland system with the efficient computational power of the extreme learning machine to achieve accurate GNSS timing data prediction. It is suitable for enhancing the resilience of positioning, navigation, and timing systems in critical infrastructure scenarios, ensuring the continuous operation of the system even when GNSS signals are interfered with or fail.

The invention relates to the field of metro station air conditioning system energy consumption prediction, and discloses a self-correcting metro station air conditioning energy consumption simulation system and method.The simulation system comprises a main cause analysis module and a self-correcting module, and the self-correcting module establishes a corresponding energy consumption simulation model and optimizes the energy consumption simulation model; the main cause analysis module comprises a data acquisition module and a data analysis module; the self-correction module comprises a model building module and an optimization module. The invention further discloses a simulation method based on the system, principal component analysis is performed on measured data of the subway station air conditioner system, an optimization algorithm is utilized to automatically correct an energy consumption simulation model, and the corrected energy consumption simulation model is applied to the subway station air conditioner control system. Therefore, the energy efficiency of the air-conditioning system of the subway station is quickly and accurately predicted, and efficient and low-consumption operation of the air-conditioning system is improved.

The application relates to the field of complex assembly process mechanical modeling and prediction, in particular to a kind of assembly process mechanical response curve prediction method and system fusing physical consistency constraint, the method comprises the following steps: constructing structure parameter space, obtaining assembly force-displacement curve dataset under different parameter combinations by finite element simulation;Resample and normalize the data, construct training samples with structure parameters and displacement as input and assembly force as output. Construct a mechanical response curve prediction model to establish a nonlinear mapping from input to assembly force. Identify the peak section and the stable section of the end section in the curve, construct a physical consistency area alignment loss, and constrain the change trend of the prediction curve in the key area to be consistent with the real curve. Combine the global fitting loss and the physical loss to construct a joint loss function to train the model. Input the to-be-predicted parameters and each displacement point into the model point by point to reconstruct the complete assembly force curve. Efficient and physically credible assembly force curve prediction is realized.

The application provides a rotary kiln ring thickness prediction system and method based on a CFD simulation technology, and relates to the technical field of rotary kiln ring thickness prediction.The application simulates the overall combustion state of the interior, the diffusion state of the interior flue gas and the heat transfer condition of the rotary kiln body when the rotary kiln is in operation through the CFD simulation technology to obtain rotary kiln inner wall temperature distribution data;according to the obtained rotary kiln geometric and operating parameter information, combined with the rotary kiln body heat transfer model, a large number of simulation and calculation of multiple working conditions, multiple time points and variable parameters are carried out to establish a variable working condition rotary kiln ring prediction database, and the growth condition of the rotary kiln ring under different working conditions and at different time points is realized in real time on site.The application realizes accurate prediction of the ring thickness when the rotary kiln equipment is in operation, avoids manual kiln shutdown and improves the production capacity.

The present application relates to the technical field of medium and shallow coalbed methane development, in particular to a kind of medium and shallow coalbed methane well production dynamic prediction method, comprising the following steps: S1, target well data acquisition and preprocessing;S2, random forest algorithm filters production main control factor;S3, fusion attention mechanism's CNN-LSTM multimodal time series feature extraction;S4, physical constraint deep learning model construction;S5, mixed loss function construction and model training;S6, production dynamic prediction and result output.The present application fuses physical constraint and deep learning technology, both utilize the efficient feature extraction capability of deep learning, and also guarantee the physical consistency of prediction result by physical constraint, solve the problem that pure data driven model generalization ability is poor, long-term prediction error is big.

The application discloses a method for optimizing process parameters of metal type nuclear fuel preparation process, collecting and arranging metal type nuclear fuel preparation process parameters and performance parameters, post-preparation performance parameters and microstructure metallographic pictures; establishing a metal type nuclear fuel solidification phase field model according to a solidification thermodynamic path under different process parameters; solving the metal type nuclear fuel solidification phase field model by using a Fourier spectrum method to obtain a phase field variable numerical solution; post-processing and visualizing the phase field variable numerical solution to generate a microstructure distribution map of the metal type nuclear fuel after solidification, and comparing and verifying the phase field model with the metal type nuclear fuel microstructure metallographic picture; calculating the post-solidification performance parameters of the metal fuel based on the microstructure distribution of the metal type nuclear fuel after solidification; taking the preparation process parameters as input and the performance parameters as output, constructing a nonlinear mapping proxy model between the metal type nuclear fuel preparation process parameters and the performance parameters; and solving the optimal preparation process parameters by using the nonlinear mapping proxy model given the target performance parameters of the metal type nuclear fuel.

The invention discloses a high-proportion new energy short-circuit current prediction method based on a GWOGAC-IELMsin mixed model. The invention relates to a high-proportion new energy power grid short-circuit current intelligent prediction method. The method comprises the following steps: constructing a hybrid prediction model GWOGAC-IELMsin based on an improved grey wolf optimizer and an incremental extreme learning machine; fault transient signal features are rapidly extracted in a 0.2 ms time window through wavelet transform, a high-dimensional input vector containing a fault phase angle, a voltage and current instantaneous value and a change rate thereof, d / q axis current reflecting new energy power electronic equipment features and other features is constructed, and key features are screened through grey correlation analysis; optimizing an input weight and a hidden layer offset parameter of the incremental extreme learning machine by adopting an improved grey wolf optimization algorithm; and finally, dynamically adjusting the network structure through an incremental learning mechanism until the prediction precision requirement is met. The method effectively solves the three technical problems that in high-proportion new energy power grid short-circuit current prediction, a traditional model is insufficient in transient feature capture, an optimization algorithm is prone to local optimization, and prediction speed and precision are difficult to balance.

The invention discloses an attention learning mechanism-based inverter model control prediction method, system and device, and a storage medium, and the method comprises the steps: pre-constructing a three-phase inverter simulation model, carrying out the real-time sampling of a state signal through employing an MPC controller, and obtaining a signal characteristic set; performing feature selection on the signal feature set, and extracting feature data; setting a screening threshold to screen the feature data, and taking the screened feature data as optimal features; inputting the optimal feature into a pre-trained machine learning model module to obtain a prediction result of the control signal of the three-phase inverter; the control performance can be improved while more calculation amount and communication traffic are reduced.

The application discloses a kind of infinite long cylinder group target strength calculation method of arbitrary arrangement, comprising: establishing cylinder group, the radius and center coordinates of each cylinder are determined;Determine the first-order scattering expression of each cylinder, and carry out local coordinate system conversion, the first-order scattering expression of each cylinder after conversion is superimposed, and the sound field coupling relationship between each cylinder in cylinder group is obtained;Rigid boundary condition is applied to each cylinder surface in cylinder group, and the first-order scattering coefficient of single cylinder is solved;On the basis of considering multiple scattering effect between cylinders, the first-order scattering coefficient of single cylinder is used to recursively obtain the multi-order scattering coefficient of cylinder group;The scattering sound field of each order of cylinder group is calculated;Each order scattering sound field is superimposed, and the total scattering sound field of cylinder group is obtained;The target strength of cylinder group is calculated.The application can solve the problem that the efficiency of underwater cylinder group structure target strength prediction is lower and modeling is complex in the prior art.

The application provides a model training method, a protein processing method and device, an electronic device, a computer readable storage medium and a computer program product. The model training method comprises: determining a first amino acid sample sequence of a first protein sample, the first amino acid sample sequence being an amino acid sequence of the first protein sample before mutation; fusing a first preset variant type and the first amino acid sample sequence to obtain a first fusion sequence; predicting the first fusion sequence by using a to-be-trained protein prediction model to obtain a second amino acid sample sequence, the second amino acid sample sequence being an amino acid sequence included in a protein variant of the first preset variant type; and adjusting parameters of the to-be-trained protein prediction model based on a first loss value of the second amino acid sample sequence and a preset third amino acid sample sequence to obtain a trained protein prediction model. Through the application, the acquisition efficiency of the protein variant can be improved.

The invention relates to the technical field of metal structure corrosion resistance data processing, in particular to a simulation calculation evaluation method for corrosion resistance of a multi-welding-seam area, which comprises the following steps: establishing a three-dimensional finite element model according to welding seam geometry and material parameters of a base metal / heat affected zone / weld nugget area; directly importing the obtained continuous residual equivalent stress field into a three-dimensional finite element model in the form of an interpolation function or an external field, and constructing stress dependent electrochemical parameters in an electrode dynamic boundary condition; after parameter setting is completed, electrode surface corrosion potential distribution and local current density are solved; and evaluating the corrosion resistance of the welding seam area according to the current density distribution, the corrosion rate field and the time-varying corrosion depth data obtained by simulation. According to the scheme, the method can be used for judging the sensitive position of the multi-welding-seam area, the corrosion resistance of the multi-welding-seam area can be evaluated more systematically and visually, the possible corrosion sensitive position can be recognized in advance, and risks can be found more timely in engineering application.

The invention relates to the technical field of animal breeding, in particular to an SNP (Single Nucleotide Polymorphism) molecular marker and application thereof in identifying pheasant laying number traits. On the basis of a genome version ASM 414374v1, the SNP molecular marker comprises any one or more of the following components: NW022205446.1: 5714958, NW022205446.1: 5715290, NW022205446.1: 5726623, and NW022205446.1: 5743232, and the SNP molecular marker comprises any one or more of the following components: NW022205446.1: 5743232. Four SNP molecular markers related to the pheasant laying number character are obtained through research and screening, and prediction and identification of the pheasant laying number can be achieved by detecting the polymorphism of the SNP molecular markers. The SNP molecular marker provided by the invention can be used for improving the laying number character of pheasants, and has important application value.

This invention relates to a rapid, non-destructive testing method and system for lipid content and deterioration degree in red pine nuts based on hyperspectral imaging and deep learning. The invention belongs to the field of non-destructive testing technology for agricultural and forestry product quality and safety. It addresses the problems of existing methods for detecting lipid content and oxidation degree in red pine nut kernels, which generally suffer from complex operation procedures, high costs for large-scale testing, long processing times, and difficulty in achieving full-process testing during storage and transportation. Key technical points: The original near-infrared spectral data of pine nut samples are obtained through hyperspectral acquisition; the true values ​​of lipid reference content and reference degree of oxidation deterioration of pine nut samples at different sampling times are determined, and a database is established based on these values. The original near-infrared spectral data of red pine nuts are preprocessed, and the data from different batches of red pine nut samples are divided into training and validation sets. An improved one-dimensional dilated convolutional network based on a dynamic weight allocation module is designed to construct a deep learning model suitable for the spectral feature analysis of red pine nut kernels. The constructed model is trained using a backpropagation algorithm, and the network weight parameters are updated in reverse by minimizing the loss function. When the constructed model meets the requirements for rapid detection, it can be used for efficient and non-destructive simultaneous detection of lipid content and oxidative rancidity in red pine nuts.

Provided is a traffic safety assistance system capable of improving traffic safety, convenience, and smoothness of a plurality of traffic participants in an object traffic area. The traffic safety assistance system includes mobile terminals that move together with people or traffic participants as mobile bodies in the object traffic area, and a cooperative assistance device capable of communicating with the mobile terminals. The mobile terminals include notification devices that perform risk notification in a care notification mode or a simulation notification mode. The cooperative assistance device includes an object traffic area recognition unit that recognizes recognition targets including each traffic participant and a traffic environment in the object traffic area, and acquires recognition information; a prediction unit that predicts future risks of a plurality of traffic participants in a partial monitoring area of the object traffic area as prediction targets; and a risk notification setting unit that sets an action mode of risk notification for each assistance target based on the recognition information and the prediction results.

The application discloses a high-rise building cluster typhoon resistance resilience prediction method based on a double-flow convolutional neural network, and aims to solve the problem that the existing high-rise building cluster typhoon resistance resilience prediction method based on physical simulation cannot meet the large-scale rapid prediction demand.The prediction method comprises the following steps: 1, a typhoon time series is combined with a virtual high-rise building cluster to obtain a virtual typhoon disaster scene, and the typhoon resistance resilience indexes of each building in the virtual high-rise building cluster are calculated; 2, the preprocessed typhoon time series and the building cluster space configuration are taken as input data, and a pixel-level label image of the typhoon resistance resilience indexes is taken as output data to construct a data set; 3, a double-flow convolutional neural network model is constructed; and 4, the double-flow convolutional neural network model is trained.Through the technical path of "simulated generation of a data set-double-flow network fusion of space-time features-end-to-end mapping of output resilience distribution", the high-rise building cluster typhoon resistance resilience under typhoon disasters is rapidly and accurately predicted.

The invention provides a superplastic forming part thickness distribution prediction method based on an agent model. The method mainly solves the problems that finite element calculation is low in speed, the requirement for the computing power of a computer is high, the calculation precision highly depends on accurate model establishment and boundary condition definition, and a prediction result cannot be rapidly migrated to other similar working conditions. According to the scheme, the method comprises the steps that parameterized representation is conducted on the shape of a part, and design interval design is conducted; extracting a plurality of groups of sample points in the design space by using a Latin hypercube sampling method; performing superplastic forming simulation on the sample points by using a finite element calculation method, and extracting grid node coordinates and thickness to establish a data set; establishing a part thickness distribution prediction agent model by taking part geometric parameters and grid node coordinates as input and part grid node thickness as output; and based on the proxy model, establishing a part thickness distribution prediction cloud picture by using the grid node coordinates and the prediction thickness value. According to the method, superplastic forming thickness distribution prediction can be carried out on the part, the prediction precision is high, the calculation efficiency is high, and the part thickness distribution prediction cloud picture can be quickly generated according to input of geometric parameters. Besides, in order to solve the problem that the local error of the part fillet area is large, the invention provides a weighted loss function based on geometric features, and the problem of the local error is improved by improving the contribution proportion of the fillet part in the loss function and influencing the training direction of the neural network.

The invention provides an air-cooled motor winding temperature prediction method and device, relates to the technical field of electric driving, and is used for predicting the temperature of an air-cooled motor winding in scenes such as electric aircrafts and electric automobiles. The method mainly comprises the following steps: modeling a motor winding temperature of each rotating speed stable section by using an index approaching model, and solving undetermined parameters of the index approaching model; based on a convective heat transfer formula, the heat dissipation coefficient of each rotating speed stable section is calculated through the motor heat loss, the environment temperature and the steady-state temperature; fitting the steady-state temperature, the time constant and the heat dissipation coefficient under different powers by using a polynomial fitting function to obtain the steady-state temperature, the time constant and the heat dissipation coefficient under a given power; and correcting the steady-state temperature of the given power according to the new environment temperature, and predicting the motor winding temperature under the given power through the corrected steady-state temperature of the given power by using an exponential approaching model.

The application discloses a coal mine goaf gas enrichment area prediction method based on an elliptical throw zone theory. The method collects goaf geological and mining parameters, and carries out standardized pretreatment. A deep learning model based on LSTM is constructed, and its gating mechanism is used to capture the nonlinear time sequence relationship between the geological and mining parameters and the gas enrichment area distribution. The elliptical throw zone theory is introduced, and the elliptical throw zone shape characteristics formed by the movement of the overburden strata of the goaf are taken as prior knowledge, which is fused with the LSTM output to realize accurate prediction of the spatial position of the gas enrichment area. Finally, a two-dimensional and three-dimensional distribution map of the gas enrichment area is generated through a visualization technology, which provides a scientific basis for goaf gas extraction and disaster prevention. In this way, the defects of low prediction accuracy and poor adaptability of traditional methods under complex geological conditions are overcome, and efficient and adaptive prediction of the gas enrichment area is realized.

The invention relates to the technical field of artificial intelligence, and particularly discloses a splashing lubricating oil distribution prediction method based on geometric deep learning, which comprises the following steps: acquiring multi-working-condition simulation data to generate a training sample, constructing a geometric deep learning model, encoding and inputting to obtain encoding features, operating the encoding features through a processor, fusing the encoding features with condition features, and predicting the distribution of splashing lubricating oil. And inputting into a decoder to predict the stable position of the oil particles. Through the constructed geometric deep learning model, after input characteristics of oil particles to be measured are input, an oil distribution result can be rapidly output through the geometric deep learning model, the stable position of splashing lubricating oil can be predicted, and the stable position of the oil particles in a real physical space of a gear box can be obtained through reverse normalization. And rapid prediction of oil liquid distribution in gear splash lubrication is realized.

The application discloses a kind of spoke non-pneumatic tire vertical stiffness prediction method and its application, the prediction method includes: according to stress-strain curve to determine the constitutive model of outer tread material, determine the preselected spoke material constitutive model set and strain parameter set;According to entity tire size to establish tire geometric model;The constitutive model parameters of outer tread material are assigned to the outer tread component in geometric model, traverse the parameters in preselected spoke material constitutive model set and strain parameter set, obtain multiple tire simulation models of different spoke constitutive model and different fitting strain;Multiple tire simulation models are respectively simulated, and effective tire simulation model is determined according to the simulation curve of counterforce-displacement;The thickness and the number of spoke of effective simulation model are adjusted, and the relationship of the vertical stiffness of tire and the thickness and the number of spoke is obtained: according to the relationship, the vertical stiffness of non-pneumatic entity tire of different spoke thickness, spoke quantity is predicted.

The application discloses a kind of based on Fourier neural operator's ground surface flood rapid calculation method, and proposes a kind of fusion two-dimensional water dynamic model and physical constraint Fourier neural operator's urban flood modeling and numerical calculation method.The method is introduced multi-scale feature mapping and physical constraint mechanism in the framework of traditional shallow water equation to solve the process of ground surface flood evolution, realize the nonlinear mapping from low-resolution simulation results and multi-source environmental driving data to high-resolution flood response field, so that under the premise of guaranteeing physical law constraint, the application realizes the high-precision, rapid prediction of urban flood process.

The invention belongs to the technical field of engines, and particularly relates to a diesel engine in-cylinder heat release curve real-time reconstruction method based on injection parameter variation, which comprises the following steps of: extracting diesel engine structure, control and operation parameters from a steady-state engine characteristic database, determining a typical load working condition and obtaining a reference combustion heat release rate curve; the circulating air inflow is calculated based on the rotating speed of the diesel engine and the pressure of an air inlet main pipe, the in-cylinder excess air coefficient is calculated according to the current circulating air inflow and the oil injection amount, the injection duration and the combustion duration are calculated in combination with the rotating speed of the diesel engine, the injection pressure, the circulating oil injection amount and the excess air coefficient, and the reference combustion heat release rate is further calculated. And phase translation is carried out on the reconstructed combustion heat release rate curve according to the main injection advance angle, and a target combustion heat release rate curve consistent with the main injection moment is obtained. The method is small in calculation amount and high in reconstruction speed, can adapt to the influence of injection parameter change on the combustion heat release characteristic under the transient working condition, and meets the application requirements of real-time analysis and vehicle-mounted control.

The application discloses a kind of oil and gas production prediction method and device.Therein, the method includes: obtaining the characteristic data of the core collected from target reservoir, obtaining the percolation data in target reservoir according to the core measured, obtaining the oil-bearing data in target reservoir according to the core measured, engineering data for oil and gas exploitation of target reservoir;The oil and gas production in target reservoir is predicted by spectral clustering model according to characteristic data, percolation data, oil-bearing data and engineering data, wherein the spectral clustering model is obtained by model training according to sample core corresponding sample data, and sample data at least includes: sample core corresponding sample characteristic data, sample percolation data, sample oil-bearing data, sample engineering data, sample oil and gas production data.The application solves the technical problems that the prediction of oil and gas production in related technologies is time-consuming and high-cost.

The application provides a rocket engine unstable combustion oscillation characteristic diagnosis method and device, relates to the rocket engine test field, and the method comprises the following steps: obtaining component parameters of each component of a thrust chamber of a rocket engine to be analyzed; determining nozzle outlet acoustic admittance based on nozzle parameters and channel parameters, and determining mass flow pulsation of each nozzle channel outlet; processing engine combustion chamber design parameters and configured flame parameters to determine single flame heat release pulsation; based on the configured modal cutoff value, the mass flow pulsation and the heat release pulsation, the configured acoustic characteristic network is solved to obtain a plurality of actual main frequencies and modal numbers corresponding to each actual main frequency, and the modal category of each actual main frequency is determined. According to the dynamic response characteristics of the injection pulsation of a single nozzle or multiple nozzles and the heat release pulsation of the combustion flame, the actual main frequency of the low-temperature engine when oscillation combustion occurs under different working conditions can be predicted, and relevant parameter analysis can be carried out.

The invention relates to the technical field of biological information, in particular to a system and a method for predicting drug sensitivity of generic cancer brain metastatic tumor and a medium. According to the drug sensitivity prediction system for the generic cancer brain metastasis tumor, the sensitivity of a brain metastasis sample to a drug is predicted by constructing a key regulatory factor set of the brain metastasis tumor sample. The drug sensitivity prediction system for the pan-cancer brain metastatic tumor can carry out drug sensitivity prediction and screening on a brain metastatic tumor sample in a targeted mode, operability is high, prediction can be completed only through an RNA expression profile, and butt joint with a clinical conventional sequencing process is facilitated; the output of the prediction result of the prediction system is based on the key regulation factor set, the interpretability is high, the accuracy is high, and low-cost, rapid and interpretable screening of drugs can be realized. The drug sensitivity prediction system for the generic cancer brain metastasis tumor provides key technical support for clinical drug screening of the generic cancer brain metastasis tumor.

The invention relates to an actuating unit excited bearing rotor position and attitude prediction and calculation method, which is characterized in that gas film mechanics numerical solution is taken as a physical basis, a prediction model is constructed in combination with a multi-input multi-output neural network, and the precision of the prediction model is verified on a test set by using indexes such as multiple correlation coefficients and the like. According to the method, the parameters of the actuating unit can be quickly predicted and reversely solved, so that the rotor position attitude is quickly and reversely solved, the active regulation and control of the rotor position attitude are realized by changing the excitation parameters of the actuating unit and dynamically adjusting the distribution of a gas film pressure field, the time delay can be remarkably reduced, and the control real-time performance and stability are improved; moreover, the method can cover various bearings and different actuation excitation modes, gives consideration to the solving speed, the application range and the convenience, and is higher in practical value.