6results about How to "Eliminate performance impact" patented technology

The invention discloses a uniform circular array robust underwater target azimuth tracking method and system, which adopts a P-element uniform circular array with a radius of r to receive narrowband acoustic signals radiated by a target; each sensor on the uniform circular array converts the received underwater acoustic signal into an electrical signal, sets the time step k=0, initializes the target state estimation value, the root mean square error matrix, the degree of freedom estimation value, and the inverse scale matrix estimation value; calculates the target state estimation value of the kth time step according to the one-step predicted target state of the kth time step, the one-step predicted root mean square error matrix of the kth time step, the degree of freedom estimation value, and the inverse scale matrix estimation value; takes the first target azimuth estimation value of the target state estimation value of the kth time step as the underwater target azimuth tracking result of the kth time step, sets the time step k=k+1, continuously tracks the underwater target azimuth, and completes the uniform circular array robust underwater target azimuth tracking. High-precision, robust and efficient target azimuth tracking is achieved.

The invention discloses a preparation method of a steel wire rope for testing. The preparation method is specifically implemented according to the following steps: step 1, preparing a lead-bismuth alloy; 2, a steel wire rope with the length not smaller than 1 m is cut out; step 3, removing the head of the cut steel wire rope; 4, grease on the detached rope head is cleaned, and moisture is removed in an air-drying mode; and 5, the lead bismuth alloy is used for casting the rope head. According to the preparation method of the steel wire rope for testing, the steel wire rope head can be manufactured by a hot casting method, the reliable consolidation of a casting material to the steel wire rope is ensured, the casting temperature is effectively reduced to 200 DEG C or below, the influence of the thermal action on the performance of the steel wire rope is eliminated, the casting material can be more efficiently recycled, and the energy consumption is effectively reduced.

The application is specifically a network security identification method and system based on multi-modal fusion features, relates to the technical field of network security, and comprises: a flow reorganization and sequence extraction module, a multi-dimensional entropy feature calculation module, a phase space reconstruction and trajectory analysis module, and a topological geometry judgment module.In the application, the adjustable parameter Renyi entropy can accurately capture the content confusion degree difference of different types of encrypted traffic, the box dimension fractal dimension quantifies the self-similarity of the traffic, and the maximum Lyapunov index is combined to realize accurate mining of time sequence chaotic characteristics; and the flow reorganization and sequence extraction module supports multiple network access modes, adapts to different bandwidth scenes, and rearranges out-of-order packets to guarantee data authenticity.

The application provides a high-speed high-precision analog-digital converter and a performance improvement method of the analog-digital converter, the analog-digital converter comprises an input configuration module, an analog-digital conversion module, an adaptive parameter extraction module and a whole cycle data recovery module, in a parameter extraction working mode, the modified parameters are extracted through the cooperation among the input configuration module, the analog-digital conversion module, the adaptive parameter extraction module and the whole cycle data recovery module; in a normal working mode, the digital calibration and correction of the digital signal are realized through the analog-digital conversion module and the modified parameters, the modified parameters comprise a jitter correction parameter, a gain correction parameter and a mismatch error correction parameter, the signal-to-noise ratio and the linearity of the analog-digital converter are improved, the influence of the performance of the analog-digital converter caused by the smaller and smaller intrinsic gain and the larger and larger mismatch error with the reduction of the process size is avoided, the injected random jitter is completely eliminated, and the influence of the signal-to-noise ratio of the analog-digital converter caused by the injected random jitter is avoided.

The invention discloses a signal detection method and a Laplacian correlation detector solving device and method. The signal detection method comprises the following steps: acquiring to-be-detected signals under two channels; solving a preset Laplacian correlation detector according to the to-be-detected signal to obtain a detection statistic; determining a detection result according to a comparison result of the detection statistical magnitude and a preset threshold value; wherein the preset threshold value is calculated through a preset false alarm probability. The signal detection efficiency can be improved.

The invention relates to the technical field of distributed artificial intelligence and collaborative perception, and discloses an intelligent perception and decision-making system for distributed multi-modal data fusion, which comprises the following steps: acquiring multi-field real-time data streams in a live broadcast process, aligning the multi-field real-time data streams according to a unified time granularity, and storing the aligned data streams in a sliding window buffer area; carrying out abnormal point detection on the deviation degree sequence of each field by applying an accumulation and detection algorithm, and outputting abnormal alarm information when the accumulation deviation exceeds a dynamic threshold value; outputting the type pre-judgment of the current exception in the field of triggering the exception alarm; when a trend type abnormal event is detected, calculating the occurrence probability and urgency degree of each downstream event, and generating early warning information and intervention suggestions; according to the invention, the pre-judgment from the current domain exception to the future associated domain event is realized. The problem of pre-judgment lag caused by single-field detection is solved, and a time window is won for operation intervention.