11results about How to "Improve engineering realizability" patented technology

The invention provides an F-P spectral imaging system and method, and belongs to the technical field of spectral component measurement, and the system comprises a front lens which is used for collecting light energy and converting the light energy into emergent light meeting the limitation of the clear aperture of an F-P interferometer; the annular band optical filter comprises a central area and a plurality of concentric annular areas surrounding the central area, the bandwidth of each area is the same as the free spectral range of the F-P interferometer, and the annular band optical filter is used for eliminating level aliasing in emergent light of the front lens; the F-P interferometer is used for carrying out wavelength selection on emergent light of the annular band optical filter and outputting an interference field; the imaging mirror is used for projecting the interference field to the detector; and the detector is used for converting the projected optical signal into a digital electric signal. The F-P spectral imaging system which can maintain high spectral resolution and also can realize wide spectrum coverage is constructed by introducing the annular band optical filter which is accurately matched with the FSR of the F-P interferometer.

This invention relates to a multi-field-of-view coding method and system based on intermediate image plane polarization mask modulation, belonging to the field of space optics technology. It addresses the technical problems in existing technologies where feature aliasing and mutual interference easily occur when multi-field-of-view signals are superimposed, and where encoded features are easily distorted in the presence of optical errors or system deviations. The coding method includes the following steps: multi-field-of-view division; polarization coding parameter allocation; polarization modulation implementation; optical folding and signal superposition; and coding result output. The multi-field-of-view coding method and system based on intermediate image plane polarization mask modulation of this invention, by introducing polarization modulation in the imaging stage, maps multi-field-of-view spatial information to the polarization dimension, realizing an extension from traditional "scalar intensity coding" to "vector polarization coding."

This invention discloses a heuristic dynamic memory weighting and recall reordering method, comprising six steps: initial impression value calculation, Ebbinghaus forgetting curve modeling, recall depth parameterization, multi-dimensional independent sorting, multi-dimensional result fusion, and recall-triggered impression value inversely proportional increment. This invention simulates the natural decay of memory through an 8-stage Ebbinghaus forgetting curve, achieves a recall-as-reinforcement effect through an inverse impression value increment mechanism, comprehensively considers semantic similarity, temporal freshness, and memory importance through multi-dimensional reordering, and flexibly adjusts the breadth and depth of recall through recall depth parameterization. This invention provides two forgetting curve implementation methods: discrete-stage lookup table method and continuous curve calculation method (including exponential decay type and power-law decay type), suitable for different application scenarios. This invention significantly improves the intelligence level and anthropomorphism of AI long-term memory systems, enabling AI to perform natural memory recall like humans.

This invention discloses a three-dimensional breast ABUS image classification method, electronic device, and computer-readable storage medium based on a tokenized bi-branch selective state-space model. The method preprocesses the three-dimensional ABUS volume data and inputs it into a hierarchical pyramidal classification network. The network constructs local convolutional branches and a global state-space branch in its basic modules: the local branch uses lightweight three-dimensional grouped convolution to extract texture and boundary morphology; the global branch aggregates features into a token map through voxel token generation, then performs multi-axis bi-directional selective state-space scanning and adaptive fusion using routing weights. After detoxing, the core features are injected through a gating mechanism. Finally, the classification result is output through three-dimensional global average pooling and a fully connected layer, enhancing the ability to model long-range dependencies across slices with near-linear complexity.

The invention belongs to the technical field of flow control, and discloses a jet flow control system for inhibiting reverse pressure gradient curved surface flow separation, which is characterized in that through an embedded multi-cavity jet structure, additive manufacturing integral forming / modular replacement and pneumatic modulation and communication control modes, on the premise that a complex mechanical reciprocating mechanism is not adopted, the reverse pressure gradient curved surface flow separation is inhibited, and the reverse pressure gradient curved surface flow separation is inhibited. And uniform output of jet flow of a nozzle, quick change of a jet angle and convenient setting of jet flow intensity and pulse parameters are realized, so that the engineering realizability of the device, the working condition switching efficiency and the repeatability of control input are improved.

This invention discloses a method and system for optimal efficiency collaborative control of SOFC-GT systems based on real-time optimization. It employs a two-layer architecture combining upper-layer real-time optimization and lower-layer staged tracking control: the upper layer periodically solves a constrained optimization problem based on the system efficiency model, real-time load demand, and equipment status, outputting the target power of the solid oxide fuel cell and gas turbine that maximizes the overall system efficiency; the lower layer receives the target power command and dynamically adjusts actuators such as fuel flow and air flow through a staged control strategy, including a fast response stage based on PI control and a steady-state optimization stage based on PID control, and introduces key efficiency parameters as auxiliary criteria for stage switching. This invention can balance system efficiency, dynamic response speed, and operational safety under varying load conditions, improving the feasibility of collaborative control of SOFC-GT systems.

The application discloses a contact structure with active vacuum arc dispersing function based on internal phase change driving, which comprises a static side contact piece, a static side conductive rod fixedly connected with the static side contact piece, a dynamic side contact piece oppositely arranged with the static side contact piece, and a dynamic side conductive rod fixedly connected with the dynamic side contact piece, wherein arc dispersing devices are arranged in the static side contact piece and the dynamic side contact piece, and the two arc dispersing devices are symmetrically arranged with the radial center line of the contact structure as the symmetric axis. The arc dispersing device is a hollow sealing structure, the side close to the contact surface is a flexible metal diaphragm, the inside is filled with liquid low-boiling point working medium, and the bottom is provided with a capillary core structure made of porous sintered material. The contact structure can drive the phase change of the built-in working medium by the energy of the arc itself, generate transient mechanical pulse to make the arc root continuously move on the anode surface, thereby destroy the stable heat conduction of the arc to the local anode, inhibit the anode ablation from the source, and improve the anti-ablation ability and service life of the vacuum switch.

The invention discloses an active power distribution network security domain description method considering inertia and voltage support, which comprises the following steps: firstly, establishing a branch load flow model of an active power distribution network, then establishing an energy storage system operation model considering virtual inertia support capability, secondly, establishing an operation model considering voltage-reactive V-Q droop control of the photovoltaic inverter, wherein the model comprises droop control constraint and photovoltaic capacity constraint; and finally, proposing an adaptive piecewise linear reconstruction algorithm, and efficiently and accurately describing a P-Q security domain of interaction between the active power distribution network and the main network, which meets the inertia requirement and the voltage reactive power control, by adaptively adjusting the sampling step length, and further constructing a P-Q-inertia three-dimensional security domain which meets the system inertia requirement and the voltage reactive power control constraint.

This invention discloses a unified five-modal target detection method, relating to the fields of computer vision and multimodal target detection technology. This method targets visible light, infrared, synthetic aperture radar, multispectral, and hyperspectral images. First, it performs unified annotation and three-channel input unification processing on different modal inputs. Then, it constructs a unified target detection network comprising five modality-specific input adapters, a shared backbone network, a shared feature fusion network, and five modality-specific detection heads. Based on the modality identifier of the input sample, the corresponding input adapter and detection head are selected to achieve modality-specific input correction, shared feature extraction, and modality-specific detection output. During the training phase, a round-robin single-batch single-modality joint training mechanism is adopted, and the loss function of the detection head corresponding to the current modality is calculated only. During the inference phase, the target detection result for the corresponding modality is output. This invention can achieve target detection of five heterogeneous modal images within a unified framework, possessing good versatility and engineering application value.

The invention discloses an airborne high squint SAR imaging method based on two-dimensional wave number spectrum resampling, and the method comprises the steps: constructing an azimuth ambiguity resolution function, correcting an aliasing azimuth wave number spectrum generated under a PRF limited condition, and providing complete two-dimensional wave number spectrum data for subsequent processing. According to the method, an azimuth ambiguity resolution function is utilized, an aliasing azimuth wave number spectrum is corrected, a two-dimensional wave number spectrum resampling factor is used in subsequent operation, a squint two-dimensional wave number spectrum is subjected to two-dimensional resampling and equivalently converted into a forward side-looking wave number domain signal, and the utilization rate of the wave number spectrum is increased. According to the method, the aliasing azimuth spectrum can be corrected under the condition that the PRF is limited on the airborne large squint SAR, two-dimensional high-resolution imaging of the target area is achieved through the imaging method of two-dimensional wave number spectrum resampling, and the engineering application scene of the method is widened.

The present application relates to the technical field of robot motion control, and especially relates to a kind of quadruped robot reinforcement learning motion controller training, simulation, deployment method based on elevation map, the method constructs the asymmetric Actor-Critic deep reinforcement learning model based on double-estimator hybrid explicit and implicit feature state estimation, and the body state estimator and the terrain state estimator are respectively extracted body perception implicit feature and foot end terrain perception implicit feature;Build multiple simulation training terrains and design course teaching mechanism and multiple types of reward functions for training;Respectively build simulation verification environment based on MuJoCo and Gazebo, construct local elevation map by ray detection or laser radar point cloud fusion, and splice as the strategy network input with the body state of robot, the present application effectively improves the motion stability, environmental adaptability and simulation to reality migration reliability of quadruped robot in complex terrain.