51results about How to "Implement refactoring" patented technology

The invention discloses a rapid three-dimensional reconstruction method for a large converter, and belongs to the technical field of point cloud three-dimensional reconstruction. The method comprises the following steps: S1, data acquisition: scanning different parts of a buggy ladle, acquiring point cloud data of each part of the furnace top, the furnace body and the furnace bottom of the converter, and performing data preprocessing; s2, feature point detection: carrying out feature point detection on the converter point cloud data preprocessed in the step S1; s3, point cloud segmentation and point cloud registration are carried out on the point cloud data obtained through preprocessing; and S4, performing three-dimensional reconstruction: performing curved surface reconstruction on the converter point cloud by using a three-dimensional reconstruction algorithm to obtain a converter three-dimensional model. According to the method, the features of point cloud three-dimensional reconstruction are utilized, more comprehensive and accurate overall three-dimensional information of the converter is obtained, and detailed geometric information of the converter can be provided.

The invention discloses a buried hill core porosity calculation method based on a high-precision repair CT image, and the method comprises the steps: S1, obtaining a micron CT three-dimensional gray image of a buried hill fracture reservoir plunger sample in a research region, and obtaining a three-dimensional digital core model of the buried hill fracture reservoir plunger sample based on the micron CT three-dimensional gray image; s2, obtaining nanometer CT three-dimensional grayscale images of the sub-samples at different positions; s3, establishing a buried hill fractured reservoir core image database under different resolutions, learning and training the image through a residual network model to obtain a high-precision repaired micron CT image, and simulating and calculating the porosity through threshold segmentation; and S4, determining the actually measured porosity of the buried hill crack sample through an indoor experiment, and comparing the actually measured porosity with the simulated calculation porosity. According to the method, the pixel boundary contact information of the micron CT image is recovered through the deep residual network model, the restoration quality of the image is improved, and the simulated porosity is closer to an actual value.

The invention discloses a non-contact electrocardiosignal monitoring reconstruction method and system based on time-frequency domain combination, and relates to the technical field of radar signal processing, and the method comprises the following steps: obtaining a chest wall vibration signal S of a to-be-detected person by adopting a frequency-modulated continuous wave radar; separating a heart mechanical motion signal SH from the chest wall vibration signal S; and converting the heart mechanical motion signal SH into an electrocardiosignal ECG through a pre-established ResNet architecture neural network model. Time domain loss and frequency domain loss are used for training a ResNet architecture neural network model, and the error between the calculated heart rate and the real heart rate is quantified through the mean absolute error and the root-mean-square error. According to the method, the neural network capable of being used for ECG signal conversion is designed based on the ResNet architecture, after the neural network is subjected to data training, more detailed signal features can be learned, the fitting accuracy is improved, then the ECG waveform can be accurately predicted, and non-contact electrocardiosignal monitoring and reconstruction are achieved.

The invention belongs to the technical field of fault diagnosis, and particularly relates to a gearbox fault diagnosis method and system based on cyclic pulse characteristic decomposition, and the method comprises the steps: constructing a new dictionary matrix, namely a cyclic pulse dictionary matrix, to accurately estimate a pulse cycle period contained in a signal formed by mixing a plurality of periodic pulse signals; a Goertzel algorithm is adopted to obtain corresponding frequency amplitudes according to the magnitude sequence of the energy values of the pulse cycle periods; and separating and reconstructing the CPCC corresponding to the pulse cycle period, and decomposing the signal into a plurality of CPCCs. According to the invention, the cyclic pulse dictionary matrix is constructed to carry out sparse representation on the pulse cycle period of the periodic pulse signal, and the cycle period of the fault pulse characteristic in the gear vibration signal is accurately estimated; the Goertzel algorithm is adopted to decompose and reconstruct the CPCC in the gear vibration signal according to the pulse cycle period, and the decomposed CPCC contains abundant periodic pulse characteristic information.

The invention discloses an SAR interference intelligent reconstruction method and system based on a complex signal domain coding and decoding network, and the method comprises the steps: S1, constructing an SAR interference intelligent reconstruction model based on the complex signal domain coding and decoding network, wherein the model comprises an input preprocessing module, a coding feature extraction module, an improved connection module, a decoding feature fusion module and an image restoration module; s2, constructing a data set for training a model; s3, training the SAR interference intelligent reconstruction model based on the complex signal domain coding and decoding network by adopting the training set, and storing a trained weight parameter; s4, pre-stored weight parameters are loaded, and a trained SAR interference intelligent reconstruction model based on the complex signal domain coding and decoding network is obtained; and S5, inputting an SAR echo signal to the trained SAR interference intelligent reconstruction model based on the complex signal domain coding and decoding network for interference reconstruction. According to the invention, extraction, reconstruction and cancellation of interference signals in SAR signals are realized.

The invention discloses an SPH-WENO numerical simulation method applied to a solid strong impact problem, and relates to the technical field of SPH numerical simulation for an elastic-plastic material strong impact process, and the method comprises the steps: carrying out the high-precision WENO reconstruction of the left and right states of a contact particle pair based on an SPH frame, carrying out the interpolation of a missing template point in a WENO template through second-order Taylor expansion, and carrying out the calculation of the WENO. Reconstructing a left state and a right state of the particle pair; and constructing a coordinate rotation-independent contact solution frame, substituting the left state and the right state obtained by reconstruction into the frame structure for iterative solution, and obtaining the contact solution of each component of the physical parameter of the particle. Therefore, by adopting the SPH-WENO numerical simulation method applied to the solid strong impact problem, a WENO high-order format under an SPH framework can be constructed, and left and right states including a density velocity vector and a stress vector can be obtained, so that a high-precision contact solution is obtained, and low-dissipation and high-precision numerical simulation for the solid strong impact problem is realized.

The application discloses a photoelastic transparent soil three-dimensional stress-deformation full-field cooperative measurement device and a use method thereof. The device comprises a loading system, a stress-strain measurement system and a model test system. The model test system comprises an optical platform and a model groove arranged at a central position of the optical platform and used for containing a sample to be tested. The loading system is arranged above the optical platform and used for applying axial pressure to the sample to be tested. The stress-strain measurement system comprises a camera module, a laser module, a light source module, a guide rail and a control system II. During work, the camera module, the laser module, the light source module and the guide rail are adjusted by the control system II, so that the camera is scanned around the sample to be tested, deformation images are collected, full-field strain is obtained, and a three-dimensional digital model of the soil body is reconstructed. The device has high integration, can obtain stress distribution information of any cross section and any direction in the material, and overcomes the inherent defects that a two-dimensional photoelasticity cannot reflect stress gradient in the thickness direction.

The invention discloses a method for directional regulation of starch-based edible film performance, and belongs to the technical field of food packaging. The invention prepares edible film by enzymatically modifying starch, and uses starch branching enzyme to hydrolyze starch and transglycosylate, and regulates α, 1-4 and α, 1-6 glycosidic bonds in the starch molecular structure, thereby realizing the preparation of highly branched starch-based film. Through the transformation of enzyme action process conditions, the structural basis and molecular mechanism of highly branched starch film formation are explored, and the performance optimization and dynamic regulation of edible film are realized, which lays a foundation for the application of starch film in different food fields, and provides new ideas and new means for food environmentally friendly and degradable packaging and application.

The application discloses a reconstruction method and system for a cavity under a shallow reinforcement net shelter, and comprises the following steps: constructing a layered medium scene containing a shallow reinforcement net and a cavity; acquiring image data information with different shelter states and preprocessing to construct a reconstruction data set; constructing a reconstruction initial model for the cavity under the shallow reinforcement net shelter and training to obtain a reconstruction model for the cavity under the shallow reinforcement net shelter; and using the obtained reconstruction model for the cavity under the shallow reinforcement net shelter to perform actual reconstruction for the cavity under the shallow reinforcement net shelter. The application acquires multiple types of training data through the constructed layered medium scene containing the shallow reinforcement net and the cavity, and trains the reconstruction model for the cavity under the shallow reinforcement net shelter which comprises a U-shaped network, a Sobel operator, an attention mechanism, a residual structure, an average mechanism and a skip connection scheme. Therefore, the application can not only realize the reconstruction for the cavity under the shallow reinforcement net shelter, but also has higher reliability, better accuracy and higher reconstruction efficiency.

The invention relates to a radiation meter dense feed source array directional diagram optimization method based on a generative adversarial network. The method comprises the following steps: constructing a training data set, a condition vector containing a target directional diagram performance index and a corresponding optimal amplitude-phase weight vector; designing a conditional generative adversarial network model; constructing a physical constraint loss function by embedding a differentiable electromagnetic calculation module; loading the constructed training data set, and training a complete generative adversarial network model based on a loss function embedded with a differentiable electromagnetic calculation module; and when the system antenna pattern obtained according to the amplitude-phase adjustment parameters output by the generative adversarial network model meets the antenna pattern evaluation index, completing dense feed source array antenna pattern optimization. According to the method, the target weight is dynamically adjusted and optimized on orbit, so that a single model can adapt to beam forming requirements in different observation scenes, and flexible switching of multiple task modes such as high-resolution detection and wide-range general survey is realized.

The invention discloses a fault reconstruction method of a digital micro-grid, and particularly relates to a fault reconstruction method of a digital micro-grid considering uncertain renewable energy power generation and a demand side resource aggregator. According to the method, a load aggregator demand response model based on a price elastic coefficient is established, and the uncertainty of renewable energy power generation accessed into the novel power distribution network is effectively described through chance constraint. In addition, a power flow model suitable for novel power distribution network reconstruction is established, and then a new energy power distribution system reconstruction model is established by considering network loss and switching cost. Aiming at the non-convexity of the reconstruction model, a large M method and second-order cone relaxation are adopted, and an intermediate variable is introduced to convert the reconstruction model into a mixed integer second-order cone programming problem. According to the method, through simulation experiment analysis, detailed verification results are given, and the reliability and effectiveness of the method in the aspect of fault reconstruction of the new energy power distribution system with demand side resources and uncertain renewable energy power generation are explained.

The invention discloses electronic equipment and a method for controlling an antenna. The electronic equipment comprises a feed source; the first antenna comprises a first radiation branch knot and a second radiation branch knot; a feeding point is arranged between the first end and the second end of the first radiation branch knot, a first grounding point is arranged between the feeding point and the first end, a second grounding point is arranged between the first grounding point and the first end, the feeding point is used for being connected with a feed source, and the first grounding point is used for being connected with a first switching circuit and being grounded; a third grounding point is arranged between the third end and the fourth end of the second radiation branch knot, a fourth grounding point is arranged between the third grounding point and the fourth end, the third grounding point is used for being connected with the second switch circuit and grounded, and an electric coupling gap is formed between the first end and the third end; when the first switch circuit and the second switch circuit are in an off state or an on state, the first antenna generates a first radiation pattern or a second radiation pattern. Wherein the main radiation directions of the first radiation pattern and the second radiation pattern are different.

The invention provides a current reconstruction method for a permanent magnet motor driving system, which is used for solving the problem of fault-tolerant operation when a current sensor breaks down due to strong electromagnetic interference on a permanent magnet motor for a multi-rotor unmanned aerial vehicle. According to the method, a first-order sliding-mode observer is designed for observing current, and on the premise that estimated current and actual current are assumed to have the same phase, a current reconstruction algorithm is designed by using an error between a measured value and an observed value of alpha-phase current on a static coordinate system, and # xa7b5 orthogonal to the alpha-phase current is obtained; the error between the measured value and the observed value of the phase current is further substituted into the designed first-order sliding mode observer to realize current observation, so that the target of reconstructing the B-phase current and the C-phase current by using the sampled A-phase current is achieved; therefore, an accurate current feedback signal is provided for current fault-tolerant control of the permanent magnet motor when the current sensor fails under strong electromagnetic interference.

The present invention belongs to the technical field of rare earth permanent magnet materials, and specifically relates to a high-abundance rare earth rapid quenching alloy and a preparation method thereof. In order to synergistically improve the solubility of La, the Ce 3+ proportion and magnetic properties in the high-abundance rare earth rapid quenching alloy, the high-abundance rare earth rapid quenching alloy of the present invention is expressed in atomic percentage as (La 1.53 Ce 3.57 RE 11.9 )-M x Fe 78‑x -B6, where 0 < x < 1; RE is an optionally added rare earth element selected from one or more of Nd, Pr, Dy, and Tb; M is a doping element selected from at least one of Ag, Sn, Au, and Cu, and the mixing formation enthalpy of the binary system formed by M and Fe is greater than the mixing formation enthalpy of the binary system formed by La and Fe, and it is prepared by a two-step method of vacuum induction melting and melt rapid quenching. The present invention can significantly improve the solubility of La in the matrix phase, and while maintaining a high proportion of Ce 3+ , it greatly optimizes the coercivity and the maximum magnetic energy product.

The application provides a kind of floating wind turbine physical-numerical mixed scale model test method, belongs to offshore wind power generation technical field, the test method will floating wind turbine modeling into two parts of physical model and numerical model, in addition to physical model and numerical model, test platform also includes measurement system and loading system;Real-time interaction is carried out between the measured signal of measurement system and the calculation result of numerical model by computer, signal loading is realized by loading system, and the construction and closed-loop control of hybrid test platform are realized.Finally, the required information is monitored and extracted through the visual interface.The application can not only accurately capture the performance of the dynamic response of the key structure, but also solve a series of problems such as scale effect, incompatibility of scale rule caused by scale effect and difficulty in simulating indoor test of offshore floating wind turbine.

The present invention discloses a spin wave transmission channel reconstruction method, a computer device and a storage medium. The spin wave transmission channel reconstruction method includes injecting current into a magnon waveguide, controlling the current density and duration of the current, and when the Bloch domain wall moves to the target position, removing the current to allow the Bloch domain wall to relax freely. The present invention can control the position of the Bloch domain wall in the magnon waveguide by injecting current into the magnon waveguide with the Bloch domain wall, thereby reconstructing the spin wave channel and realizing the control of the spin wave transmission path by current; since the current injected into the magnon waveguide directly acts on the device itself, no additional wiring is required, the complexity of the device used is low, and the spin wave channel is controlled by injecting current, so it is suitable for nano devices, so that it is easy to obtain good performance such as high integration brought by small-size devices. The present invention is widely used in the field of electronic device technology.

The present application belongs to the technical field of biomaterial surface modification, and particularly relates to a multi-level micro-nano structure mixed phase titanate hydrogen release coating and a preparation method and application thereof. Nanometer titanium oxide and calcium carbonate are used as raw materials, and the two are uniformly mixed through a mechanical ball milling method. PVA is added for bonding and granulation, grinding and sieving to obtain a powder that can be sprayed. First, an initial coating containing titanium / calcium is prepared on a pure titanium substrate through plasma spraying, and the prepared coating is subjected to alkaline hydrothermal treatment to in-situ grow a mixed phase titanate micro-nano sheet structure. On this basis, the coating is further subjected to wet steaming treatment to destroy and reduce the micro-nano sheet, and in-situ induce the growth of dense secondary nano particles on the surface of the micro-nano sheet, so that the coating has better ammonia borane loading capacity, realizes stable and long-acting hydrogen release for more than 15 days, can effectively eliminate oxidative stress in a complex environment, and has excellent biocompatibility.

The present invention belongs to the technical field of new energy solid oxide fuel cells, and specifically relates to a recyclable PCFC cathode material and a preparation method and a recycling method thereof. The present invention aims at the technical problems that the existing proton ceramic fuel cell (PCFC) cathode material has poor matching with common electrolyte materials, is easy to detach, causes battery performance degradation, and lacks a complete PCFC cathode recycling technology. The present invention provides a new type of recyclable PCFC cathode material and a preparation method and a recycling method thereof. The new type of PCFC cathode material has the advantages of high activity and recyclability. The recycling method is simple to operate, has low energy consumption, and the recovery rate can theoretically reach 100%, which is easy to promote and use.

The present invention provides a reconfigurable flexible pressure sensor based on magnetic ion droplets and a preparation method thereof, wherein superparamagnetic nanoparticles are combined with ion droplets to prepare magnetic ion droplets with high fluidity and magnetic controllability, and the magnetic ion droplets are used as the pressure sensitive layer of the pressure sensor device to construct a flexible stretchable pressure sensor based on the magnetic ion droplet array. The present invention also provides a distributed / combined non-contact multi-element in-situ reversible reconstruction method of the flexible pressure sensor, which uses an external magnetic field to control the magnetic ion droplets in the sensor to perform programmable low-resistance rapid movement and deformation between the super-hydrophobic interconnected microfluidic channels and chamber arrays, so as to regulate the spatial distribution, composition / volume and geometric shape changes of the magnetic ion droplets in the array, and achieve the distributed / combined non-contact multi-element in-situ reversible reconstruction of the performance of the flexible stretchable pressure sensor in terms of spatial resolution, sensitivity, detection limit and pressure response mode through one or more combinations of changes.

The present invention relates to the field of cable cover plate detection technology, and discloses a cable cover plate anomaly detection method and system based on a variational autoencoder, including denoising the original data of the cable cover plate collected by a sensor to obtain a denoised signal, and dynamically resampling the denoised signal according to the local change rate to obtain resampled data; inputting the resampled data into a pre-established data reconstruction model to obtain reconstructed data, the data reconstruction model is constructed based on a variational autoencoder and trained based on a dynamic learning rate adjustment strategy and a data enhancement strategy; performing anomaly detection according to the reconstruction error change rate of the reconstructed data to obtain abnormal data, and generating corresponding alarm reminders according to the abnormal data. The present invention can not only improve the integrity and reliability of data, improve the flexibility and accuracy of anomaly detection, but also can respond to changes in the operating status of the cable cover plate in a timely manner, and improve the real-time and effectiveness of anomaly identification.

The invention discloses an industrial control private network protocol vulnerability mining method based on protocol analysis. The method comprises the steps of 1, preprocessing industrial control flow data; 2, extracting a traffic fingerprint in the traffic data set; and step 3, analyzing the specification of the target protocol by using the traffic fingerprint. And 4, reconstructing a state transition track of the target protocol by using the analysis specification. And step 5, using the state conversion track to guide the test case to test the state of the tested target, and at the same time, using a heartbeat message detection mode to monitor the state of the tested target and position the triggered vulnerability. The invention further provides an industrial control private network protocol vulnerability mining system based on protocol analysis. According to the method, the analysis accuracy of the private protocol specification can be effectively improved, and meanwhile, more protocol states can be subjected to efficient fuzzy testing under the condition of the same number of test cases, so that more vulnerabilities hidden in the protocol are mined.

The invention discloses an extremely weak light time domain information extraction method and device based on single photon detection, and belongs to the technical field of optical communication and quantum time frequency transmission. In order to solve the problems that link attenuation is high and far-end extremely weak light is difficult to detect in non-relay amplification long-distance time transfer, a single-photon detection technology is mainly adopted to be combined with a scanning chromatography technology, clock signals are synchronized through a time base module, a light source module, a modulation module, an amplification module, an optical fiber transmission module and a single-photon detection module are matched, and the time base module and the single-photon detection module are combined. Reconstruction of extremely weak light waveforms and extraction of key time domain information are realized. According to the invention, time-frequency information can be effectively transmitted in long-distance optical communication without relay amplification, and the detection difficulty of extremely weak light is overcome.

The present invention discloses an SPH-WENO numerical simulation method applied to the problem of strong impact of solids, and relates to the technical field of SPH numerical simulation for the strong impact process of elastoplastic materials, including high-precision WENO reconstruction of the left and right states of the contact particle pair based on the SPH framework, and interpolation of the missing template points in the WENO template using the second-order Taylor expansion to reconstruct the left and right states of the particle pair; constructing a coordinate rotation-independent contact solution framework, substituting the reconstructed left and right states into the framework structure for iterative solution, and obtaining the contact solution of each component of the physical parameters of the particle itself. Therefore, by adopting the above-mentioned SPH-WENO numerical simulation method applied to the problem of strong impact of solids, it is possible to construct a WENO high-order format under the SPH framework, obtain the left and right states containing the density velocity vector and the stress vector, thereby obtaining a high-precision contact solution, and realizing low dissipation and high-precision numerical simulation for the problem of strong impact of solids.

This application relates to a water purification device, a control method for the water purification device, an apparatus, a computer device, and a medium. The water purification device includes: a multi-stage filtration assembly connected between an inlet and an outlet, with the filtration level of each filtration assembly increasing sequentially from the inlet to the outlet, and each filtration assembly being connected to subsequent filtration assemblies via pipelines; multiple valve assemblies installed on the connecting pipelines of each filtration assembly; a water quality detection assembly installed between the first-stage filtration assembly and the inlet; and a controller used to determine at least one target filtration assembly from among the filtration assemblies based on the water quality status reflected by water quality information, determine the valve adjustment parameters of each valve assembly based on the target filtration assembly, and control each valve assembly to adjust the water flow direction according to the valve adjustment parameters, thereby using the target filtration assembly to filter the raw water. This method can improve the service life of the purification equipment and reduce operating costs.

This invention relates to a negative electrode current collector with a composite aperture array, its preparation method, and its application. The negative electrode current collector includes a copper foil substrate on which a composite aperture array structure is provided. The composite aperture array structure includes multiple through holes and multiple blind holes. The through holes are arranged through the thickness direction of the copper foil substrate to provide lithium ion transport channels. The blind holes are arranged along the thickness direction of the copper foil substrate and are closed at the bottom to provide nucleation sites for lithium deposition and buffer space for volume expansion. The through holes and blind holes are alternately distributed according to a preset arrangement rule. In any local area, at least one through hole and multiple blind holes constitute a combined unit to form a current density difference distribution on the surface of the copper foil substrate and guide lithium deposition. The preset arrangement rule includes: the through holes and blind holes are arranged periodically in a plane according to the combined units. The combined unit includes at least one through hole located at the center and multiple blind holes arranged in a ring around the through hole.

The present application relates to the field of biomedical engineering, specifically provides a kind of self-powered water-volt battery dressing with self-release water and water function, the dressing innovationally integrates temperature-sensitive conductive hydrogel layer and water-volt power generation layer, realizes the self-adaptive, self-sustaining intervention to wound microenvironment, through the unique "water storage-release-water" closed-loop mechanism, system utilizes body temperature trigger water release on demand, and realizes the self-supply of fuel and energy recycling by absorbing tissue exudate and external water environment moisture, so as to overcome the technical bottleneck that the energy source of existing self-powered system is unstable, intermittent supply, ensure that continuous, stable therapeutic electric stimulation is provided in the whole healing period;Compared with traditional treatment methods, the dressing can significantly accelerate wound closure and promote ordered and high-quality regeneration repair.

The invention provides a reconfigurable power amplitude limiter and a multi-stage reconfigurable power amplitude limiter, and the reconfigurable power amplitude limiter comprises a power detection unit which is selectively connected with a radio frequency port through a plurality of mutually independent power detection ports so as to detect input power and generate control voltage according to the input power; and the power discharge unit is selectively connected with the radio frequency port through a plurality of mutually independent power discharge ports and limits the amplitude of the output power of the radio frequency port under the action of the control voltage. The power amplitude limiter can be reconstructed by selectively connecting the power detection port and the power discharge port, and different amplitude limiting output requirements are met.

The present application relates to sensing technology and seabed topography monitoring field, aims at providing a kind of test device and method for analyzing the coupling law of sensing array and soil deformation.The device includes the main box of top opening and the scanning mechanism arranged at the opening, side push mechanism and vertical push mechanism are arranged in the main box;Variable volume test cavity is formed by deformable side plate and multiple side push oil cylinders arranged in side push mechanism, and vertical push bottom plate and multiple side push oil cylinders arranged in vertical push mechanism, for containing test soil and burying sensing array.The present application uses telescopic oil cylinder and push plate combined mechanism to simulate soil settlement, lifting and collapse, which can realize the simulation of soil in multiple directions, different displacement deformation variables and deformation speed, and the main device is simple to process and low in material cost.The deformation coupling law of sensing array with different scales and different burial depths, different soil, different scales, different deformation variables can be studied.

The invention discloses an incremental grid reconstruction method based on adaptive parameters, and the method comprises the steps: firstly obtaining point cloud data, carrying out the linear point cloud segmentation of an obtained point cloud based on a point index, dividing the point cloud into m subsets, and enabling the adjacent subsets to have an overlapping region; then density estimation is carried out on the segmented point cloud subsets, and downsampling and normal estimation are carried out; and finally, constructing a triangular mesh for each point cloud subset by adopting an incremental method. According to the method, through point cloud segmentation and overlapping region introduction, the problem of computing resource limitation is solved, and the consistency and continuity of a processing result are ensured through the design of the overlapping region; the dynamic adjustment of down-sampling and normal estimation is realized through adaptive parameters obtained by density estimation, so that the retention of important geometric features is ensured, and the overall processing efficiency is improved; dynamic vertex combination and patch reconstruction are realized by defining a vertex mapping function and a distance threshold.

The invention discloses a pulse signal reconstruction method based on deep learning in a non-Gaussian environment, and relates to a pulse signal reconstruction method based on deep learning. The invention provides a pulse signal reconstruction method based on deep learning in a non-Gaussian environment, and aims to solve the problems that the existing marine environmental noise often has a non-Gaussian statistical characteristic, and a traditional method based on Gaussian distribution hypothesis is low in pulse signal reconstruction accuracy under the background of non-Gaussian interference and low input signal to interference plus noise ratio. The pulse signal reconstruction method based on deep learning in the non-Gaussian environment comprises the following specific steps: constructing a training set; constructing a deep neural network model; training the constructed deep neural network model based on the training set to obtain a trained deep neural network model; and inputting the received underwater sound pulse signal into the trained deep neural network model, and outputting a reconstructed pulse signal by the trained deep neural network model. The method is applied to the field of pulse signal reconstruction.