114 results about "Differential algorithm" patented technology

The invention discloses a multi-objective optimization method and system for dual active bridge circuit expansion phase shift modulation, and the method comprises the steps: calculating a voltage matching ratio, and determining an external phase shift angle ratio based on a difference value between an output voltage and an output voltage reference value; according to the voltage matching ratio and the external phase shift angle ratio, obtaining an internal phase shift angle ratio of the primary side full bridge through an offline lookup table; calculating the maximum transmission power of the dual-active bridge converter, and determining a current working mode from the six preset working modes based on the size relationship between the inner phase shift angle ratio and the outer phase shift angle ratio; generating an optimal modulation strategy through a multi-objective optimization algorithm by taking the current effective value and the current peak value as optimization objectives and combining the voltage matching ratio and the periodic parameters of the current working mode; and generating a trigger signal of each switch tube of the dual-active bridge converter. Through six-mode working area division, complete differential algorithm boundary constraint and off-line-on-line collaborative optimization, the contradiction between the current stress and the soft switching range of the converter under extended phase shift modulation is solved.

The invention discloses a pavement temperature field parameter inversion and data fusion method, and the method comprises the steps: building a pavement temperature field solving partial differential equation based on a heat conduction theory; representing an initial condition, a boundary condition and a measuring point temperature of a pavement temperature field based on an interpolation algorithm; normalizing the solution domain, generating sampling point coordinates at the pavement temperature solution domain and the boundary based on a random sampling method, and taking the sampling point coordinates as input variables of the deep neural network; based on the coordinates of the sampling points, a normalized operator value of the sampling points is obtained in combination with a deep neural network and an automatic differential algorithm; combining initial conditions and boundary conditions, fusing actually measured temperature data, and constructing a total loss function of the deep neural network through a weighted summation method; and based on gradient descent and a total loss function, deep neural network training is carried out, adaptive sampling is carried out according to partial differential equation loss of the current neural network, and finally pavement temperature field prediction is realized. According to the invention, high-precision prediction of the pavement temperature field is realized.

The invention discloses a dam termite image recognition method based on multi-modal fusion and dynamic feature enhancement. The method comprises the steps that image data, environmental parameters and vibration signals of a dam monitoring area are collected in real time through a multi-modal sensor network; carrying out the preprocessing of the image data, and carrying out the target detection through an improved ACP-YOLOv5s model; processing the continuous video frames by adopting Gaussian background modeling and a difference algorithm to generate a binary mask graph representing dynamic activities of termites, and separating a moving target area from a static background; and the target detection result and the dynamic mask graph are fused, the collected environment parameters and vibration signals are combined, the termite damage level is calculated through an AI engine, and graded early warning or automatic killing response is triggered. The method breaks through the limitation of traditional single-mode detection. And a multi-dimensional hazard grade evaluation model is established, so that the automation level of dam termite prevention and control is greatly improved.

The invention discloses a regulation and control system and method based on heat energy of an energy storage cabinet, and belongs to the technical field of intelligent heat management. Temperature data in the energy storage cabinet are collected, a temperature distribution matrix is constructed, and a thermal gradient vector and an overheating risk area are extracted based on a difference algorithm; further establishing a thermal power load mapping model, and dynamically generating an air circulation path optimization diagram; the opening angle of a flow deflector and the rotating speed of a fan are adjusted in combination with an optimized path, and self-adaptive regulation and control of a thermal field in the cabinet body are achieved; meanwhile, a machine learning model is trained by utilizing historical operation data, an optimal regulation and control strategy is output, and an execution layer is driven to realize collaborative response of a fan and a flow deflector; the method has the advantages of high regulation and control precision, high response speed, low energy consumption and the like, and is suitable for heat management requirements of various types of energy storage cabinets.

The present invention relates to the fields of PID control and Kalman filtering, and disclosed is an improved PID control method based on Kalman filtering. The present invention comprises: in the process of acquiring a target value of an expected signal, taking the difference between the target value and a current feedback quantity of a controlled object to obtain an error value, acquiring an output of PID from the error value, applying the output of PID to the controlled object to obtain a noisy output signal, then, after Kalman filtering, feeding the output back to PID, continuously iterating to make the feedback quantity approach the target value, and finally obtaining a stable output signal. Variable-speed integration and integral separation algorithms are added to an integral term of PID to change an integration speed on the basis of a system deviation, and an incomplete differential algorithm and a low-pass filter are added to a differential term of PID to prevent high-frequency interference. According to the present invention, the optimal estimation of a system state is obtained by means of prediction of a current state and updating of observation data, thereby reducing the influence of measurement errors and noise, and improving the performance and stability of a PID controller.

The invention relates to the technical field of debugging systems, in particular to an adaptive debugging system and method for touch screen FW parameters, and the system comprises a signal acquisition module, a trend triggering module, a temperature compensation calculation module, a window generation module and a parameter output module. According to the invention, the edge of the driving voltage is carefully collected, the sliding difference algorithm is introduced to filter out high-frequency noise, meanwhile, real-time sampling of environment temperature and baseline capacitance adjustment are combined, the influence of external variables on signals can be dynamically corrected, and accurate judgment of capacitance fluctuation is improved by utilizing periodic delay deviation and regression trend analysis. By means of main frequency feature extraction and zero crossing point sudden change detection, threshold setting can be refined, self-adaptive window expansion can be achieved, touch signal response is kept sensitive and consistent all the time, adjustment and optimization can still be automatically completed even under the condition that the temperature changes frequently or signal interference is complex, parameter drift and debugging failure are effectively prevented, and user experience is improved. And the parameter automatic matching efficiency and the finished product consistency in batch production are improved.

The invention relates to a neuron magnetic stimulation numerical simulation calculation method, belongs to the technical field of crossing of biomedical engineering and neuroscience, and aims to solve the problems of lack of three-dimensional dynamic electrified physiological characteristic characterization, low stimulation parameter debugging efficiency based on experience trial-and-error and low simulation accuracy. And the problem of insufficient neural response prediction accuracy caused by transmembrane current continuous differential solution difficulty is solved. According to the method, head MRI data and neuromorphological information are fused to construct a personalized three-dimensional model, and electric field distribution of a target area is obtained through electromagnetic field simulation; the electric field intensity is decomposed along the trend of neurons, discrete segments are established, a transmembrane current difference algorithm is proposed by using the relation between adjacent nodes and the electric field gradient, and differential solution is converted into parallel difference operation; and membrane potential change is simulated by combining a Hodgkin-Hertz equation. The method constructs a personalized model, accurately obtains electric field distribution, simplifies operation by using a difference algorithm, simulates potential by combining an equation, and realizes quantitative analysis and curative effect extraction of the whole process of magnetic stimulation.

The invention relates to the technical field of video recognition and analysis, and discloses a video recognition and analysis method based on physical characteristics. Video frame pixels are mapped to a two-dimensional coordinate system with the upper left corner as an original point, and mirror image expansion and median filtering are carried out on a gray level image; constructing a binary image based on a gray threshold value, and analyzing and extracting a target region by using a four-neighborhood connected domain; using neighborhood search and polar angle sorting to close the tracking contour, and generating equidistant re-sampling points based on Euclidean distance and an interpolation method; the curvature of the re-sampling points is estimated through a three-point difference algorithm, and zero denominator is avoided through numerical protection; performing discrete Fourier transform on the curvature sequence to extract a frequency spectrum, and normalizing an amplitude to form a standardized feature vector; and finally, inter-frame similarity is calculated based on the feature vector, and the most similar frame is automatically retrieved. By processing unified data standards in stages, edge noise is suppressed, sampling uniformity is ensured, and feature stability and cross-frame comparability are improved.

The invention relates to the technical field of metal composite materials, in particular to a metal composite material production process dynamic simulation method based on digital twinning. Comprising the following steps: constructing a digital twin model comprising a physical entity layer, a virtual model layer and a connection layer, and establishing a Fourier neural operator model embedded with a Navier Stokes equation and a heat conduction equation constraint; real-time molten pool monitoring data are mapped into model input through a connecting layer, and advanced time step deduction is executed by using a deep learning model to predict the state of a molten pool; based on the differentiable characteristic of a physical information neural operator, calculating the gradient of a prediction state and an ideal target by using an automatic differential algorithm, generating a process parameter field through reverse iteration updating, and discretizing the process parameter field into a galvanometer control instruction; and when the risk of aggregation of the reinforced phase particles is predicted to exist, automatically generating a swing scanning strategy to realize active suppression. The method gives consideration to simulation precision and real-time performance.

The invention relates to a target detection and parameter estimation method for an OTFS (on-the-fly-switching) integrated system, which aims at solving the problems of target detection and fractional Doppler and time delay parameter estimation of the OTFS integrated system, and comprises the following steps: step 1, a sending end generates an integrated signal, step 2, a receiver receives an echo signal and processes the echo signal, and step 3, the received echo signal is sent to a receiving end; 3, a threshold graph is generated through the integer delay and Doppler detection network based on the convolutional neural network, and target detection is carried out; and 4, fractional Doppler and delay indexes are estimated. The target detection and parameter estimation method of the OTFS communication and sensing integrated system based on deep learning has the following advantages that compared with a common CFAR detection algorithm, the detection performance of the integer time delay and Doppler detection network is better. In addition, compared with a difference algorithm which only considers integer Doppler and time delay and considers fractional Doppler and time delay, the relative distance and speed estimated by the fractional estimation network are more accurate.

The invention discloses a mining area equipment increment upgrading method and system based on a difference algorithm, and the method comprises the steps: obtaining the GNSS positioning and base station fingerprint positioning of mining area equipment, carrying out the comparison verification with a preset electronic fence polygon region after preprocessing, and activating an upgrading task after the verification is passed; carrying out difference comparison on the old file and the new file by adopting an improved binary system difference algorithm through dynamic sliding window partitioning and a multi-level hash index, and compressing a difference part to generate an increment difference upgrade package; and establishing connection with mining area equipment through a network, transmitting the increment difference upgrade package to the mining area equipment by adopting a three-way handshake transmission protocol, and downloading, decompressing and upgrading the increment difference upgrade package by the mining area equipment. According to the method, directional upgrading can be realized in a complex network environment, the flow consumption can be remarkably reduced, the retransmission data volume is reduced, and the upgrading success rate of weak signal areas such as strip mine pits is improved.

A calculation method for a transformer winding considering a frequency-dependent loss, comprising: according to simulation requirements, defining key parameters calculated by using a finite-difference time-domain (FDTD) algorithm; constructing an equivalent flat wire model in an FDTD grid; calculating frequency-dependent internal impedance per unit length of a flat wire; and on the basis of a vector fitting method and a time-domain convolution calculation method, substituting the frequency-dependent internal impedance into a time-domain equation of the FDTD algorithm for updating and iterative solving, so as to obtain an electromagnetic field calculation result. In the present invention, on the basis of the vector fitting method and time-domain convolution calculation, the impact of the frequency-dependent characteristics of an electromagnetic field distributed in a conductor on an electromagnetic transient process is accurately considered during the FDTD time-domain solving, so that the calculation accuracy and efficiency of conventional FDTD algorithms and even conventional cross-scale modeling techniques can be significantly improved, thereby carrying out the targeted electromagnetic transient analysis on the reliability, safety, and predictability of transformers, and achieving the early detection of potential faults and defects, or formulating targeted maintenance plans and repair schemes.

The invention relates to the technical field of industrial data management. The data management method, system and device based on the industrial digital base and the medium are provided, the method comprises the steps that block chain evidence storage is conducted on version hash, micro-differential packet hash and an obtained operator certificate, an on-chain evidence storage block is generated, distributed storage processing is conducted on a micro-differential packet, and an off-chain differential packet storage entity is obtained; when it is detected that the data is abnormal, abnormal version positioning is carried out based on the record-level blood relationship chain, and an abnormal version sequence is determined; on the basis of a predefined inverse operation mapping relation, performing reverse playback on a micro-differential packet corresponding to the abnormal version sequence to realize historical data state recovery; and carrying out signature verification on the incremental service pack, carrying out dynamic loading processing after the verification is passed, updating a preset blood relationship extraction rule and a preset difference algorithm, and generating a hot update strategy, so as to achieve the technical effects of improving the data blood relationship tracing precision, reducing the storage redundancy overhead and enhancing the credibility of an evolution process.

The invention provides an intelligent HAST equipment adjusting method and system based on multi-source heterogeneous data and a medium, the heterogeneous data is collected in real time based on temperature, humidity and pressure sensors, the instantaneous change rate and the sliding window mean value are calculated through a difference algorithm, the optimal cooling mode is matched in combination with a preset rule base, and the optimal cooling mode is obtained. The problems of control lag and misoperation caused by the fact that traditional equipment depends on a single threshold value are solved; in the initialization stage, a modeling driver parameter template is loaded, dynamic adjustment is conducted according to the real-time change rate in the operation process, the fan rotating speed is optimized through a fuzzy PID algorithm, an ARIMA time sequence model is introduced to predict the future trend, the exhausting or cooling action is triggered in advance, and passive response is avoided; and when the parameter mean value exceeds a safety threshold value, executing emergency shutdown in a grading manner and recording an over-limit event to ensure the safety of equipment and samples.

The invention provides a process GPU-oriented real-time fault tolerance and task migration system and method, and relates to the technical field of distributed computation.The method comprises the steps that a training task state in a GPU video memory is captured, difference data is generated through an incremental difference algorithm, and the difference data is stored in a multi-level storage system after being subjected to lightweight compression; meanwhile, GPU node health indexes are collected in real time, node states are analyzed through a time sequence prediction model, and a migration process is triggered when fault risks are predicted or preemption early warning is received; during migration, the task is paused, a final increment check point is generated, check point data is quickly transmitted to a target healthy GPU node through an RDMA network, a complete context state is reconstructed, task execution is recovered, and real-time fault tolerance and seamless migration of the GPU task are achieved. According to the method, the continuous operation reliability and the resource utilization efficiency of the GPU cluster in the long-time training task are improved.

The present invention discloses a method for improving the prediction accuracy of soil iron oxide based on an amplified hyperspectral dataset. The method comprises the following steps: amplifying the hyperspectral dataset and performing multiple fractional-order differential transformations on the original hyperspectral spectrum using a fractional-order differential algorithm; then combining the fractional-order differential spectra with the original spectrum to form different amplified hyperspectral databases to amplify the amount of original spectral data; establishing a 1D-CNN neural network model for each of the different amplified hyperspectral databases, where the independent variable is the hyperspectral reflectance value and the dependent variable is the soil iron oxide content; and selecting the optimal 1D-CNN model based on the models established for the different amplified databases. The present invention effectively solves the problem of low accuracy of deep learning models under small sample conditions, improves the applicability and effectiveness of deep learning models to a certain extent, effectively improves the existing prediction accuracy of soil iron oxide based on hyperspectral data, and provides new methods and ideas for hyperspectral prediction research of other soil properties.

The invention discloses a coal mine field monitoring system convenient for demisting, particularly relates to the field of monitoring and flow prediction in a humidity fluctuation environment, is used for solving the problems of picture blurring and flow data distortion caused by high humidity sudden change, and is characterized in that distributed sensors are deployed in a coal mine tunnel environment and a multi-dimensional environment matrix is constructed; after a humidity abrupt change interval is identified by using a differential algorithm, attenuation characteristics of energy efficiency of the demisting device are captured through a multi-scale wavelet and morphological filtering means, and a limit intervention critical value is determined, and then the critical value and a visual ambiguity index of a monitoring picture are subjected to comparison and grading matching; the definition change of pictures before and after defogging intervention is dynamically quantified, finally, a self-adaptive clustering regression mechanism is combined to carry out multi-dimensional deviation correction on hierarchical definition indexes and actual measurement data of a flow sensor, a regression optimization curve is generated, the intervention opportunity of a key time period is calibrated, the defogging efficiency and the flow prediction precision can be comprehensively improved, and the defogging efficiency and the flow prediction accuracy are improved. And the reliability of overall monitoring and scheduling is enhanced.

The invention provides a power plant DCS data acquisition interface automatic configuration method and system. The method comprises the following steps: acquiring a production process diagram file from a DCS upper computer, and analyzing and extracting measuring point information based on a DCS type; generating a structured lexicographical order through lexical analysis, and screening target measurement points according to a preset rule to generate a measurement point list; adopting a Myers difference algorithm to compare the states of the measuring point lists, and dynamically updating the interface program acquisition range; and transmitting the measuring point list, the process diagram and the real-time data to a management information layer MIS across a safety area, and synchronously checking configuration consistency based on a digital twinborn model. According to the method, automatic extraction of the measuring points is realized through a regular expression and a DOM analysis technology, dynamic key encryption and an incremental updating mechanism are combined, the problems of low manual configuration efficiency, updating lag and cross-regional data islands are solved, the accuracy and synchronous real-time performance of measuring point configuration are remarkably improved, and the method is suitable for large-scale popularization and application. And safe and reliable digital support is provided for power plant production monitoring and information management.

The invention relates to the technical field of optical measurement, in particular to an optical scattering method based on an inverse problem solving algorithm and a mixed measurement method for batch samples thereof, and the method comprises the following steps: carrying out optical measurement on samples to obtain actually measured optical signals; obtaining prior structure information of the sample; constructing an initial model; analyzing the initial model by utilizing strict coupling wave analysis to generate a simulation optical signal; judging whether the difference between the generated simulation optical signal and the actually measured optical signal meets a set requirement or not; if yes, outputting a measurement result; and if not, gradient calculation is carried out on each parameter in the initial model, gradients of all the parameters are obtained by utilizing an automatic differential algorithm, then the initial model is optimized by utilizing the gradients of the parameters to obtain an optimized model, and the step of analyzing by utilizing the strict coupling waves is repeatedly executed until a set requirement is met. According to the method, more detailed contour optimization can be provided, and the calculation efficiency can be improved.

The invention discloses a parallel robot trajectory optimization method based on an improved dream optimization algorithm. The invention relates to the technical field of parallel robot trajectory optimization, and aims to obtain a smooth running trajectory of a robot in a high-speed running process and improve the motion performance of the parallel robot in the high-speed running process. The method comprises the following steps: constructing a target function according to total time and total impact required when a parallel robot operates according to a specified track; constructing constraint conditions according to the speed and acceleration operation requirements of the parallel robot; initializing a population of a dream optimization algorithm by means of a good point set and reverse learning; improving a dream sharing strategy by adopting a difference algorithm; and jumping out of the local optimization capability through the crossover variation improvement algorithm, thereby improving the quality of the solution and obtaining the optimal solution of the target function. The objective function considers the running time and impact of the robot, and the stability of the running process of the robot is guaranteed while the running speed of the robot is increased.

The application discloses a kind of data mirror persistent backup systems of power grid, it is related to electric power system technical field, including data acquisition module, data preprocessing module, difference data synchronization module, data integrity check module, data mirror module, data storage module, control and management module;Data acquisition module is used to collect the operation data of power grid in real time;Data preprocessing module carries out cleaning, formatting and compression processing to the data collected, prepares for subsequent data synchronization;Difference data synchronization module calculates and identifies the change part of data since last synchronization by difference algorithm, and only synchronizes the change part data.The application greatly reduces the data transmission amount when real-time topology reconstruction by only synchronizing the changed data, reduces the backup time at the same time, improves the efficiency of power grid data backup, optimizes the synchronization process of data, reduces the storage space waste caused by full backup, and guarantees the accuracy of backup data by hash value comparison mode.

The invention discloses a nuclear pulse signal trapezoidal forming method suitable for an FPGA, and the method comprises the steps: carrying out the high-speed digital sampling of an exponential decay original signal outputted by a detector, and obtaining a discrete digital detector signal vn; calculating the discrete digital detector signal vn by using a differential algorithm to obtain a differentiated pulse signal pn; delaying the pulse signal pn by using a digital delay line to form three delayed pulse signals, and performing synthesis operation on the pulse signal pn and the delayed pulse signals to obtain a synthesized pulse signal sequence sn; and performing integral operation on the synthesized pulse sequence signal to obtain a trapezoidal forming pulse yn. The nuclear pulse signal trapezoid forming method is simple in calculation and suitable for being implemented in the FPGA, few FPGA logic resources are used, and the calculation speed is high.

The present invention discloses an active filter protection method, comprising: obtaining the current flowing through the capacitor branch of the filter under normal circumstances; obtaining the action value of the double-ended differential protection based on the capacitor branch current; and determining whether to activate the differential protection based on the relationship between the action value and the action range. The present invention can protect the entire filter, eliminating dead zones in the protection range and operating conditions. Simultaneously, the use of a two-side differential algorithm provides stronger braking for out-of-zone faults and higher sensitivity for in-zone faults. Compared with multi-side differential protection, this method significantly improves the sensitivity and reliability of the protection, reduces the computational complexity and difficulty of setting coordination, and has wide applicability and high practicality.

The application is a target detection and parameter estimation method of a sensing integrated system, aiming to solve the target detection and fractional Doppler and time delay parameter estimation problems of an OTFS sensing integrated system, and the method comprises the following steps: step one, a sensing integrated signal is generated at the sending end; step two, the receiver receives the echo signal and processes it; step three, a threshold map is generated through an integer time delay and Doppler detection network based on a convolutional neural network to perform target detection; and step four, fractional Doppler and time delay indexes are estimated. The target detection and parameter estimation method of the OTFS sensing integrated system based on deep learning has the following advantages: compared with the commonly used CFAR detection algorithm, the detection performance of the integer time delay and Doppler detection network is better. In addition, compared with the differential algorithm considering only integer Doppler and time delay and considering fractional Doppler and time delay, the relative distance and speed estimated by the fractional estimation network are more accurate.

The application provides a power plant DCS data acquisition interface automatic configuration method and system. The method comprises the following steps: obtaining a production process diagram file from a DCS host computer, analyzing and extracting measuring point information based on a DCS type; generating a structured dictionary sequence through lexical analysis, and generating a measuring point list according to a preset rule to screen target measuring points; comparing the measuring point list state by using a Myers difference algorithm to dynamically update the interface program acquisition range; transmitting the measuring point list, the process diagram and real-time data to a management information layer (MIS) across a safety zone, and synchronously checking the consistency based on a digital twin model. The application realizes automatic extraction of measuring points by using regular expressions and DOM analysis technology, solves the problems of low efficiency of manual configuration, lagging update and cross-zone data island by combining dynamic key encryption and incremental update mechanism, significantly improves the accuracy of measuring point configuration and synchronization real-time performance, and provides safe and reliable digital support for power plant production monitoring and information management.

The invention discloses a micro-grid energy scheduling method and device based on an improved difference algorithm and a medium, and the method comprises the steps: predicting photovoltaic output power, load and charging pile demand power based on a pre-constructed algorithm model in combination with historical data, and determining the charging and discharging power of an energy storage device based on a pre-constructed energy storage power model; establishing a dual-objective function, and determining a corresponding constraint condition in combination with the photovoltaic output power, the load and charging pile demand power and the charging and discharging power of the energy storage equipment; dividing a power consumption time period, and checking whether the number of iterations of the algorithm reaches the maximum number of iterations by improving a difference algorithm so as to determine an optimal iteration result; and according to an iterative optimal result, through an energy router and a controller, controlling output of photovoltaic equipment, charging and discharging of energy storage equipment and an electricity purchasing strategy of a main power grid, and adjusting a load demand and the power of a charging pile so as to realize energy scheduling of the micro-grid.

Improved bit error correction for non-volatile memory can be implemented in multiple stages achieving improved correction capacity. As an example, bit error correction for a set of data can utilize a logic or a differential algorithm applied to one or more copies (N) of the set of data to produce a logic (or differential) output. An error correction code (ECC) can be applied to the logic (or differential) output to produce corrected data that corrects bit errors of the set of data, if any, up to a maximum for the ECC selected. An algorithm can be selected to address measured bit error rates or variations in bit error rates among binary bit states of a non-volatile memory.

The application belongs to the technical field of computers, and provides a PDF document comparison method and system based on intelligent differential merging. The method is as follows: the visual fingerprint sequences of a source PDF document and a target PDF document are calculated; the visual fingerprint sequences of the two PDF documents are compared through a Myers differential algorithm, a difference sequence is generated, all deletion operation sequence regions adjacent to insertion operation sequence regions are identified, and a candidate page pair is constructed; pre-filtering is performed on the candidate page pair, a similarity score is calculated, and based on a threshold judgment and a greedy strategy, the best match is locked, the deletion operation node and the insertion operation node are combined and replaced by a modification operation node; finally, spatial quantization index technology is used to compare internal objects, and the final comparison result is output. The application can not only accurately identify the subtle changes of the PDF document, but also reduce the comparison complexity, improve the comparison efficiency, and meet the real-time comparison needs of users.

The invention relates to an NPU firmware online upgrading method and device based on differential updating and a storage medium, and the method comprises the steps that based on a differential algorithm, a differential file generation module of an upgrading server compares a firmware version file needing to be updated with an existing firmware version file of target NPU equipment, and a differential file is generated; based on the communication protocol frame, the upgrade server transmits the differential file to the target NPU device through the network device; the NPU device receives the differential file, backs up the current firmware and updates the firmware; and after the new firmware is updated, verifying the integrity of the updated firmware and restarting the NPU equipment to activate the new firmware. The upgrading data volume can be reduced by generating and transmitting the differential file, and the upgrading time is shortened; the upgrading risk is reduced; resource consumption is reduced, and occupation of storage space and computing resources is reduced; and continuous transmission from a breakpoint is supported, and the waiting time and the maintenance cost of a user are reduced through an efficient and low-risk upgrading method.

The invention relates to the technical field of glacier environment observation, and discloses a multi-satellite glacier flow velocity environment monitoring system based on a GNSS differential algorithm, which comprises a glacier flow velocity testing module, a data recording collector module, a data analyzing and processing module and a communication transmission module, the data recording collector module stores and records the collected glacier flow velocity environmental element original data; the data analyzing and processing module analyzes and calculates the collected glacier flow velocity environmental element original data to obtain a first content package and a second content package, the first content package is the content of pictures and images, and the second content package is the content of characters and numerical values; the communication transmission module compresses, encrypts and transmits the first content packet and the second content packet; by reducing the content transmitted by the first content packet and the second content packet, the size of the transmitted data can be reduced, the data transmission time can be shortened, and the condition of network congestion can be reduced.