7 results about "Scale separation" patented technology

The application discloses an InSAR surface deformation multi-source vertical contribution analytical method based on deformation signal spectrum decomposition and belongs to the technical field of surface deformation monitoring. The steps are as follows: first, long-time SAR images are acquired, vertical deformation data and related rates and deformation variables are obtained through SBAS-InSAR technology processing; then, a plurality of auxiliary conditions are fused, regional spatial partitioning is realized through K-Means clustering; subsequently, CEEMDAN decomposition and mean threshold multi-scale reconstruction are performed on deformation signals of each partition, high-frequency items, low-frequency items and trend items are obtained; finally, ICA blind source separation and variance contribution rate analysis are performed, and shallow, medium and deep surface deformation information is inversed through directional reorganization; the K-Means clustering algorithm is adopted to perform spatial partitioning on the regional deformation field, the homogeneity and precision of subsequent signal processing are improved; and the CEEMDAN decomposition and multi-scale reconstruction method is introduced, adaptive and fine scale separation of deformation time series signals is realized; the blind source separation technology is applied to multi-scale component reorganization, and the physical attribution of the vertical position of the subsidence source is realized.

A wafer-scale method of making gallium nitride (GaN) device die is provided. In embodiments, the method includes: providing a GaN wafer including a GaN material layer, a non-crystalline substrate, and at least one etchable intermediate layer between the GaN material layer and the non-crystalline substrate; forming trenches through the GaN material layer and at least partially through the at least one etchable intermediate layer; forming one or more tether material layers on a first side of the GaN material layer and through portions of the trenches, thereby forming a set of tethers between the GaN material layer and the non-crystalline substrate; and removing the at least one etchable intermediate layer to expose a second side of the GaN material layer. In implementations, the resulting exposed second side of the GaN material layer has a surface roughness of less than 1 nanometers (nm) Root Mean Square (RMS).

A control method for vehicle magnetorheological suspension based on dual-timescale separation includes: establishing an equivalent drive circuit model with time delay; separating the equivalent drive circuit model into two timescales based on singular perturbation theory to construct a fast-changing current subsystem and a slow-changing control subsystem; designing a slow-changing control subsystem controller for the slow-changing control subsystem, calculating the reference current output by the slow-changing control subsystem, and then outputting the reference current to the fast-changing current subsystem; designing a fast-changing current subsystem controller for the fast-changing current subsystem, designing a Smith predictor to compensate for the time delay, and calculating the compensated duty cycle output by the fast-changing current subsystem by combining the reference current and the non-singular reaching law; and outputting the compensated duty cycle to the equivalent drive circuit model to achieve real-time control of the magnetorheological suspension. This invention can effectively handle the multi-source time delay problem of magnetorheological suspension in two timescales, improving vehicle ride comfort and handling stability.

The invention discloses a lightweight action recognition method and system based on a multi-scale separation topology network, and belongs to the technical field of computer vision and artificial intelligence. The technical problem to be solved by the invention is to overcome the defects that an original space-time diagram convolutional network is large in parameter quantity, high in calculation redundancy and difficult to deploy in real time in a clinical rehabilitation training environment with limited calculation resources. According to the technical scheme, the method is characterized in that a network architecture comprising a feature preprocessing layer and a plurality of cascaded lightweight graph convolution blocks is constructed; wherein the feature preprocessing layer realizes channel expansion and time sequence dimension reduction through convolution; the space graph convolution module adopts a channel separation strategy and a learnable adjacent matrix mechanism, and adaptively learns an optimal connection relationship between joint points while reducing parameters; the time convolution module adopts a multi-scale design, and integrates pooling and deformable convolution branch so as to efficiently capture multi-granularity time sequence characteristics from fine joint movement to a complete action cycle. The method has the main beneficial effects that on the premise of ensuring the action recognition precision, the model volume is remarkably compressed, and the reasoning speed is greatly improved, so that the real-time deployment of a high-precision action quality evaluation system on embedded equipment becomes possible, and the urgent demand of clinical rehabilitation training on real-time feedback is met.

This invention discloses a simulation calculation method for macro- and micro-scale coupled pollution at coal material induced airflow transfer points, comprising the following steps: S1: Constructing a three-dimensional model of an open underground transfer point and a flow field calculation grid; S2: Establishing a macro- and micro-scale separation strategy for solid materials; S3: Calculating macroscopic coal flow motion and spatial porosity based on the discrete element method; S4: Calculating fine dust diffusion based on a fluid dynamics discrete phase model; S5: Performing asynchronous coupled calculations of fluid and discrete element methods, which can avoid simulation result distortion and the problem of coupled simulation easily leading to grids exceeding 100 million and calculation divergence.

The invention provides a sea surface height abnormal data fusion method and device, and relates to the technical field of ocean data processing. According to the method, flattening processing is carried out on wide-swath two-dimensional sea surface height abnormal data, so that the wide-swath two-dimensional sea surface height abnormal data can be fused with along-rail sea surface height abnormal data under a unified framework; the method comprises the following steps: flattening one-dimensional data, performing scale separation on the flattened one-dimensional data, fusing different scale components by adopting an optimal interpolation method based on space-time, compensating a fusion result corresponding to a second scale by utilizing a trained compensation model, and finally reconstructing fusion results obtained under different scales to obtain a fusion result. Therefore, large-scale background information and sub-mesoscale detail information can be processed separately and reconstructed cooperatively, and the problem that a small-scale structure is easy to be smoothed excessively due to mutual coupling of different-scale information in an existing unified fusion mode is solved. The obtained target sea surface height anomaly fusion data can well consider the stability of a background field and the recovery capability of a fine structure.

The application discloses a simulation calculation method for macro-micro coupling pollution of a coal mine material induced airflow transfer point, and comprises the following steps: S1, constructing an underground open transfer point three-dimensional model and a flow field calculation grid; S2, establishing a macro-micro scale separation strategy of a solid phase material; S3, calculating macroscopic coal flow movement and spatial porosity based on a discrete element method; S4, calculating micro-fine dust diffusion based on a fluid mechanics discrete phase model; and S5, performing asynchronous coupling calculation of a fluid and a discrete element, so that the problems of simulation result distortion and grid over one hundred million and calculation divergence caused by coupling simulation can be avoided.