13results about How to "Fix performance degradation" patented technology

The application discloses a LIBS element quantitative analysis method based on double-branch feature fusion and electronic equipment, the method comprises the following steps: preprocessing the collected LIBS spectrum signal, including baseline correction and spectrum resampling; inputting the spectrum signal data after preprocessing into a double-branch feature extraction network, extracting local features through a CNN branch, and extracting global features through an MLP branch in parallel; performing adaptive weighted fusion on the local features and the global features through a gating fusion module to obtain fusion features; inputting the fusion features into a WMA-MLP model, the WMA-MLP model is based on MLP, integrates a multi-head self-attention mechanism and a residual module, is used for modeling the global dependency relationship between features, and outputs a final element quantitative analysis result. Through the parallelly arranged CNN and MLP branch feature extraction structures, more comprehensive spectrum feature representation can be obtained, and the accuracy and robustness of the spectrum analysis model are improved.

The invention relates to the technical field of copper powder preparation, and provides a preparation method of liquid-phase in-situ self-assembly nano-shell-coated micron-core composite copper powder, a specific inducer is introduced into a liquid-phase reduction system, copper sulfate is adopted as a copper salt, and sodium hypophosphite or VC is adopted as a reducing agent, so that one-time forming of a micron-core-nano-shell composite structure is realized; the problems of non-uniform particle size distribution, high energy consumption, low yield and the like in a traditional copper powder preparation method are effectively solved; according to the prepared composite copper powder, the low-temperature sintering driving force brought by the nanometer shell layer is obtained while the process-friendly characteristics of being easy to filter, easy to wash, easy to dry and the like of micron powder are reserved, the initial sintering temperature can be reduced to about 150 DEG C, and the machining performance of the copper powder under the low-temperature condition is remarkably improved.

The invention belongs to the technical field of integrated circuits, and provides a single-ended MEMS input-to-differential output circuit and an implementation method thereof. Through the CMFB circuit, effective signals of the CMEMS are averagely distributed to the VBIAS end and the VIN end, the maximum signal range which can be processed by the Pre-amplifier is doubled, meanwhile, a single-end-to-differential circuit is omitted, power consumption and noise are reduced, and circuit performance is improved. Compared with the prior art, according to the technical scheme, an impedance matching circuit or a capacitance compensation circuit is omitted, so that signal attenuation caused by parasitism is reduced, and performance is improved. Particularly, the stray capacitance of the VBIAS end can be ignored. The problem of performance deterioration under the condition that the stray capacitance of the VBIAS end is too large in the prior art is solved.

This invention discloses a surface acoustic wave (SAW) resonator, comprising a piezoelectric layer; an electrode layer on the piezoelectric layer; and a transducer comprising a first busbar, a first long finger, a first dummy finger, a second busbar, a second long finger, and a second dummy finger. The first and second busbars extend along a first direction and are positioned opposite each other. The first long finger, second dummy finger, second long finger, and first dummy finger extend along a second direction and are all located between the first and second busbars. The first long finger and second dummy finger are positioned opposite each other. The piezoelectric layer includes at least one heterostructure, the depth of which is greater than the transducer thickness. The propagation speed of SAW in the heterostructure differs from its propagation speed in the piezoelectric layer. The heterostructure extends along the first direction, and its vertical projection onto the piezoelectric layer overlaps with the vertical projection onto the piezoelectric layer but does not overlap with the vertical projections onto the first and second gaps. This invention suppresses lateral mode transmission and improves filter performance.

This invention relates to the field of electronic materials technology, specifically disclosing a copper-based varistor chip electronic paste and its preparation method. The paste is composed of a specific ratio of sheet-like and nano-copper powder, a glass binder phase, an organic carrier, and functional additives. The preparation method employs a multi-stage vacuum dynamic degassing device. This device gently releases large bubbles through a pretreatment chamber, and utilizes a dynamic stirring mechanism with combined revolution, rotation, and reciprocating motions within the main degassing chamber to achieve strong shearing and agitation of the paste across its entire surface. Combined with ultrasonic cavitation, this deeply breaks down bubbles, thereby efficiently and thoroughly removing them. Simultaneously, the device integrates an online viscosity monitoring and intelligent solvent vapor compensation system, enabling real-time adjustment and stabilization of the paste's rheological properties. This invention effectively solves the problems of incomplete bubble removal and performance fluctuations caused by solvent evaporation, significantly improving the density, uniformity, and production consistency of the paste product.

The present application relates to the technical field of sacrificial anode material, in particular to a preparation process of high-current-efficiency rare earth micro-alloyed magnesium sacrificial anode, which comprises the following steps: (1) preparing raw materials: preparing according to Mg-Al-Zn-Ce-Mn-In alloy system, controlling the content of Ce to be 0.3-0.8wt%, the content of Mn to be 0.2-0.5wt%, the content of In to be 0.05-0.15wt%, and the rest being Mg and inevitable impurities; (2) vacuum smelting and ultrasonic refining; (3) low-temperature homogenization annealing; (4) preparation of conductive type micro-arc oxidation composite coating. Through the synergistic micro-alloying design of rare earth Ce and Mn and In, the present application realizes the effective inhibition of heterogeneous precipitation of β-Mg 17 Al 12 phase and significant refinement of grains, and solves the problem of low current efficiency caused by serious micro-electric couple corrosion of traditional magnesium sacrificial anode.

The application discloses a kind of space-time domain joint's sparse drive adaptive line spectrum enhancement method, including the uniform linear array of hydrophone receives the far-field plane wave signal of signal source;The far-field plane wave signal received by entire uniform linear array is input into space-time domain joint filter;The far-field plane wave signal received by next array element is input into space-time domain joint filter as reference signal, until processing all the far-field plane wave signal received by array element, to obtain the output signal corresponding to each array element respectively;The output spectrum corresponding to each output signal is obtained respectively;Square superposition is carried out to multiple output spectrums and is averaged, to obtain output result;Output result is converted from frequency domain to time domain, obtain the input signal of norm-based sparse drive adaptive line spectrum enhancer;Input signal is input into norm-based sparse drive adaptive line spectrum enhancer and carries out weight updating;According to the length of input signal, iterative adaptive weight is continued, and corresponding output signal is obtained.

The invention discloses a performance optimization system and method for a photonic integrated device, and belongs to the technical field of photonic integration. The multi-physics field collaborative modeling module receives the device structure, the material characteristics and the process parameters, and outputs a performance prediction value analyzed by the coupling model; the process compatible optimization module receives CMOS process specifications and outputs compatible material combinations, heterogeneous integration schemes and process optimization results; the control scheduling module receives the adjustable device array parameters and outputs a time division multiplexing control signal; the optical-digital conversion module realizes bidirectional conversion between a digital electric signal and an optical phase signal; the performance closed-loop optimization module collects real-time performance data of the device, outputs optimized control parameters or structure parameters, and dynamically corrects performance deviation. By adopting the performance optimization system and method of the photonic integrated device, the photonic integrated device can efficiently cooperate with multiple module functions, the stability is improved, the service life is prolonged, the comprehensive performance is optimized, the power consumption is reduced, and meanwhile, the large-scale production cost is reduced.

The invention discloses a Bezier curve-based code element transmission waveform processing method and system, and the method comprises the steps: generating a Bezier curve based on a set of predetermined control points, and defining a vertical coordinate sample value of the Bezier curve as a code element transmission waveform; constructing a transmission message sequentially comprising a synchronization header, an information header and a data segment, wherein the information header is used for storing coordinate information of a control point set for generating the Bezier curve; modulating the transmission message and sending the transmission message to a receiving end; receiving and demodulating a code element transmission waveform signal, and analyzing coordinate information from an information header of a transmission message; restoring a Bezier curve and a corresponding code element transmission waveform by using the analyzed coordinate information; in a received data segment sampling signal, translating a restored code element transmission waveform, calculating a difference value between the waveform and a received sample value, and positioning a starting time point of each code element waveform according to a minimum difference value criterion; and resampling the data segment signal based on the positioned starting time point and demodulating original data.

This invention provides an efficiency-aware differential privacy noise injection method for fine-tuning direction-amplitude decomposition-based federated LoRA, relating to the field of data privacy protection technology. In each round of communication, the client performs a reparameterized local training phase and evaluates the local importance of LoRA updates across layers. The client prunes the LoRA parameter updates to be uploaded, introducing a hierarchical perturbation mechanism. Based on the importance level division, differentiated noise injection strategies are adopted for LoRA layers of different importance levels. Under the premise of satisfying differential privacy constraints, the redundant noise terms introduced by low-rank decomposition are eliminated from the structural level by deterministically mapping the injection method of the original noise in the weight space back to the LoRA factor space. The noise-reconstructed LoRA parameter updates are uploaded to the server for federated aggregation and broadcast back to each client as initialization parameters for the next round of local training.

The invention discloses a visual inertial navigation fused unmanned aerial vehicle trajectory tracking control method and system, and relates to the technical field of unmanned aerial vehicle navigation and control, and the method comprises the steps: obtaining multi-mode perception data, and carrying out the timestamp alignment; constructing a visual odometer network to extract a relative pose variation, and constructing an inertial integral network to extract a short-time motion estimator; a self-adaptive confidence fusion module is constructed, a visual modal weight is dynamically adjusted according to environment texture richness, an inertial navigation modal weight is dynamically adjusted according to maneuvering intensity, and self-adaptive weighted fusion is realized through a learnable attention mechanism to output high-precision pose estimation; and inputting fusion pose estimation into a nonlinear model prediction controller, and outputting an expected attitude angle and a thrust instruction to realize accurate trajectory tracking.

The invention relates to the technical field of reactive compensation of a power system, and discloses a capacitor switching fault protection method and system based on three-phase cooperative control, and the method comprises the following steps: S1, enabling an A-phase switch element, a B-phase switch element and a C-phase switch element in a three-phase circuit to be under the independent control of an intelligent control unit; s2, the intelligent control unit predefines at least two phases in the three phases as main control phases and defines the remaining phase as a standby cooperative phase; s3, when a conventional capacitor switching operation is executed, the intelligent control unit controls the standby cooperative phase to keep a closed state; s4, when the switching instruction is a disconnection instruction, starting a fault identification process; and S5, when it is judged that the fault phase exists, the intelligent control unit immediately executes collaborative protection operation. According to the invention, by setting the main control phase and the standby cooperative phase, at least two phases of circuits can still be cooperatively disconnected when a single switch fails, so that the formation condition of a resonant circuit is thoroughly blocked.

This invention relates to the field of hydrogen energy station safety technology, and particularly to a comprehensive method for detecting leaks in energy stations that integrates infrared thermal imaging and gas sensing data. The method utilizes convolutional neural networks to deeply extract the spatial features of infrared thermal imaging data and long short-term memory networks to independently model the temporal dynamic features of gas sensor data. The probability outputs of the two independent models are converted into the basic probability assignment function in Dempster-Shafer evidence theory through an uncertainty quantification method based on information entropy. Evidence reasoning and fusion are performed at the decision-making level, which not only integrates multi-source information but also explicitly models and handles the uncertainties of each piece of evidence. This application not only overcomes the shortcomings of insufficient reliability of existing single-sensor technologies and the poor fault tolerance and high model coupling of low-level data fusion methods, but also solves the problem of performance degradation of existing systems in complex environments.