274results about How to "Address usability issues" patented technology

A T-BOX information security detection and protection method based on vehicle anomaly data monitoring comprises: classifying and summarizing vehicle faults, providing a universal vehicle fault tree model, collecting and comparing vehicle real-time data through T-BOX, monitoring vehicle anomaly data in real time, transmitting to a telematics service platform (TSP) constantly, determining change trends of operation parameters and states of each vehicle part and system through data fusion and through big data analysis based on vehicle driving historical information collected by a cloud terminal and vehicle main parts related information restored by a main engine plant, comparing vehicle real-time information transmitted by T-BOX, judging whether a vehicle has anomaly data or not at present, and determining whether the vehicle is under information security attack or not based on the established vehicle fault tree model; finally, notifications or warnings on fault problems due to vehicle functional security need to be made for users and vehicles by the TSP.

The invention discloses an electric power information network vulnerability threat evaluation model establishment method faced to electric power CPS risk evaluation and an evaluation system based on the model, the model selects three groups of security vulnerability evaluation factors including vulnerability distribution threat degree, access approach and utilization complexity; an analytic hierarchy process is adopted to establish a vulnerability threat degree evaluation model, and a parameter construction method is provided; on the basis of quantitative evaluation, the evaluation result defines a vulnerability threat level, and the objectivity of quantitative evaluation and visibility of qualitative evaluation are considered; and the system related to the invention comprises a vulnerability detection part and a server part which adopt distributed structure. According to the invention, a vulnerability scanning system and a security area of the electric power information network are combined, an electric power information network vulnerability threat evaluation algorithm based on CVSS and analytic hierarchy process is adopted to perform vulnerability threat evaluation, so that the system can be safer and more accurate for performing vulnerability scanning and safety evaluation on the electric power information network.

The invention discloses a three-dimensional gesture key point detection method based on a neural network, which comprises the following steps: acquiring a gesture data set including gesture area information and gesture two-dimensional and three-dimensional key point position information; obtaining a gesture data set containing gesture area information and gesture two-dimensional and three-dimensional key point position information. Training a gesture segmentation network which can detect a gesture region in an RGB image by taking an RGB image containing a gesture as an input; the gesture region detected by the gesture segmentation network is truncated, up-sampled or down-sampled. Training a two-dimensional gesture key point detection network, which can detect a plurality of two-dimensionalgesture key points in a gesture region image; the absolute coordinates of the key points of the three-dimensional gesture are converted into relative coordinates; a 2D to 3D gesture key point mappingnetwork is trained. The network can map multiple 2D gesture key points into 3D space to form 3D gesture key points. The invention can quickly and effectively detect three-dimensional gesture key points from RGB images containing gestures.

The invention provides a mouth-shape animation generating method used in a source virtual role, which comprises the following steps: (1) a text is received and is split into individual characters, and the individual characters are split into textual phonemes; (2) the phonemes are transformed into corresponding mouth-shape commands according to a phoneme mouth-shape matching set, wherein, initial mouth-shapes and terminative mouth-shapes corresponding to the phonemes are set in the phoneme mouth-shape matching set; (3) the mouth-shapes corresponding to the textual phonemes in the mouth-shape commands are zoomed according to a phoneme duration table, wherein, the phoneme duration table comprises mouth-shape amplitude marks; (4) initial frames and terminative frames which respectively correspond to the initial mouth-shapes and the terminative mouth-shapes corresponding to the textual phonemes are calculated according to the phoneme duration table, and the execution time of the mouth-shape commands is set, wherein, the phoneme duration table comprises the corresponding mouth-shape duration of the phonemes; (5) the mouth-shape commands are executed, and a mouth-shape animation used in the source virtual role is generated. The mouth-shape animation generating method can solve the applicability problem on the aspect of mouth-shape animation copying in a RBF neural network.

The invention relates to a multifunctional modularized floating island for in-situ restoration of polluted rivers and lakes. The floating island consists of five functional modules including a waterscape floating island module, a flexible filler floating island module, an aeration floating island module, a power supply floating island module and a functional floating island module. Each functional module is formed by floating beds in the same shape and size; different functional components are matched on the floating beds respectively so as to form the above functional modules. 2-5 different functional modules are selected in proportions according to the water quality difference and area load of the polluted river, and are optimally combined, thus various micro-environments are formed along the water flow direction, the self-cleaning ability of rivers and lakes are prompted, and the COD (chemical oxygen demand) and ammonia and nitrogen of the polluted water can all reach the V-type standard of surface water. With the adoption of the technology, the problems of single function, poor applicability and low water quality improving efficiency of the plant floating island structure in the prior art are solved, and the in-situ high-efficiency water quality improvement of water in polluted rivers and lakes is achieved.

The invention provides a transcatheter aortic valve and preparation method thereof. The transcatheter aortic valve comprises a radially telescopic reticular valve frame, pedal skirts and multiple lobes, wherein the petal skirts are arranged on the inner side of the petal holder, the lobes are located in the lumen of the valve carrier, wherein the petal skirt is divided into a first region and a second region, the first region faces an inflow end of the aortic valve, the second region faces the outflow end of the aortic valve and the second region is provided with a mesh hole capable of communicating the inside and outside of the valve frame. The transcatheter aortic valve provided by the invention solves the problems that the coronary orifice is easy to be blocked after the aortic valve isimplanted in the prior art and the aortic valve is suitable for a patient.

The invention discloses a semi-automatic digital method of a non-standard map. The method comprises the following steps of: 1) scanning to obtain a grid image, and performing geographical space registration and correction and binarization; 2) establishing a linear vector layer and assigning a geographical space reference system consistent with the corrected grid image; 3) overlapping the linear vector layer on the grid image after the binarization, identifying and capturing the center lines of the lines in the grid image, automatically tracking the captured center lines of the lines in the grid image one by one, and performing line breakage and endpoint artificial judgment on each automatically tracked line, and performing vectorization to obtain a vector image; 4) performing attribute assignment, check and correction; and 5) re-establishing the topological relation of the vector image, and checking and correcting the topological errors such as overlapping and gap to obtain a vector image after digitalization. The method disclosed by the invention has the advantages of good image processing accuracy, high image processing efficiency, low processing workload, short processing period and wide application range.

The invention provides a method and a system for searching a non-structural electronic document with obvious category classification. The method comprises a document classification and type keyword recognizing stage and a real-time searching stage; in the document classification and type keyword recognizing stage, document classification is used for classifying documents in specific collection according to a relation existing among content of the documents, and type keyword recognizing is used for recognizing keywords in all types; and in the real-time searching stage, the documents which are in accordance with a searching result are searched according to searching words input by a user and returning the searching result from high to low according to a document correlation, wherein a real-time searching correlation algorithm formula is provided by the real-time searching stage, the relation between searching words and document types is introduced in the real-time searching correlation algorithm formula, a TF-IDF (Term Frequency-Inverse Document Frequency) algorithm is optimized, two problems when the TF-IDF algorithm is used for searching electronic documents of a large-scale enterprise are solved to a large extent, and thus the method and the system are suitable for the full-text search of the electronic documents of the large-scale enterprise.

The present invention relates to a chirp pulse amplifying spectrum shaping dielectric film structure reflector, which comprises a transparent substrate, a high reflectance film system, an anaglyph structure and an external protective layer. Wherein, the high reflectance film system is composed of a plurality of staggered dielectric films. A high reflectance film layer and a lower reflectance film layer of the high reflectance film closely contact the anaglyph structure. The anaglyph structure can take a plurality of shapes, such as a micro-lens high transparency film system structure, an air dielectric structure, a glass anaglyph structure or other transparent dielectric structures, including semiconductor structures. As a high-power laser chirp pulse with a plane wave structure vertically casts onto the reflector, the laser passes through the high reflectance film system and the anaglyph structure and all residual lasers are reflected to back of the reflector through the transparent substrate. Reflex intensity distribution is modulated to a needed spectrum distribution structure. The reflector of the present invention can be inserted into any place of an amplifier link and improve capacity to distinguish shaping spectrum chromatic dispersion to a certain level. Scope modulation exceeds 60% without changing phase position, thus adapting to PW devices.

The invention relates to a micro-type dynamic piezoresistive pressure sensor and its making method, using MEMS silicon bulk micromachining method to make a pressure sensitive chip of E-shaped silicon cup structure, the reverse side of the chip is welded with static seal technique, Pyres glass ring and Hitachi alloy ring, to form silicon back contact medium quasi-level packaging, the reverse back of the chip is covered with anti-interference insulating layer, the right side of the chip is connected with a lead cable through connecting circuit, the sensor tube cap with cable drawing-off mouth at the tail is hermetically welded to the Hitachi alloy ring to form sealed isolation between the side pressure surface and the back pressure cavity, the tail of the tube cap squeezes tightly the cable without sealing, the nozzle part of the sleeve covered on the cable is hermetically covered on the tail of the tube cap, the heat-shrinkable tube segment covered on the cable is hermetically fixed with the outer end part of the sleeve nozzle, thus implementing atmosphere connection between the back pressure cavity and the measured environment, and it is used in shrinkage mould test for hydraulic engineering like dam, ship gate, pier, embankment, etc. and has low measuring range, high sensitivity, strong anti-interference performance, and good dynamic performance during fluid dynamic test.

The invention relates to a civil aircraft undercarriage assembling and disassembling vehicle suitable for multiple aircraft types. The civil aircraft undercarriage assembling and disassembling vehicle consists of a mobile chassis, a ball screw lifting device, a six-degree-of-freedom parallel fine adjustment mechanism, a force sensing device and a multihole detachable support. By redesign of the civil aircraft undercarriage assembling and disassembling vehicle, translation operation at any direction is realized by McKenna's wheels; large-travel lifting operation is realized by four ball screw mechanisms; multi-orientation position and attitude fine adjustment operation is realized by the six-degree-of-freedom parallel fine adjustment mechanism; assembling and disassembling work of the undercarriage of multiple aircraft types is realized by the multihole detachable support; the force sensing device is additionally arranged at the chassis part of the six-degree-of-freedom parallel fine adjustment mechanism, so that precise force control during unloading is realized; the whole assembling and disassembling process is automatically controlled. By the use of unique module structure design, the problems of single applicative aircraft type, difficulty in alignment operation, complicated operation process, low adjustment precision, low efficiency and the like of the conventional civil aircraft undercarriage assembling and disassembling vehicle are solved.

The invention discloses a staggered MIMO radar array antenna construction method, which is a construction method of a staggered double-end emission model and a staggered inside emission model based on the phase center approximation (PCA) principle. The staggered double-end emission model can obtain bigger array antenna apertures but introduce more jitter errors caused by airfoil jitter while the staggered inside emission model can reduce the array antenna apertures, but the introduced jitter errors can be obviously reduced. The models can be flexibly selected according to different system design requirements. The method better solves the applicability problems under the bent wing condition and the array interval restriction condition while maintaining the characteristics of easy analysis and design of the linear arrays.

The invention relates to an engine stalling protection controlling method under a high altitude condition. The engine stalling protection controlling method under the high altitude condition comprises steps as below: a. measuring the atmosphere pressure of an engine; b. according to a preset relationship chart of atmosphere pressure calibrated inside an electronic control unit (ECU), an engine target idle speed rotary speed, a first stalling protection rotary speed and a second stalling protection rotary speed, determining the calibration values of the mentioned rotary speeds corresponding with the measured atmosphere pressure; c. measuring the current rotary speed of the engine; d. judging whether the current rotary speed meets the conditions set by the following steps, if the answer is yes, corresponding process is conducted, otherwise returning to step c: d1. if the current rotary speed is lower than the first stalling protection rotary speed calibration value, at least one load device which currently runs is enabled to stop working and return to step c; d2. if the current rotary speed is lower than the second stalling protection rotary speed calibration value, at least one load device which currently runs is enabled to stop working and return to step c; d3. if the current rotary speed is larger than the engine target idle speed rotary speed, at least a part of load devices which stop working due to d1 and / or d2 is recovered to normal working, and then returning to step a.

The invention relates to a rigidity variable robot towards a minimally invasive surgery and a working method. The rigidity variable robot comprises a mechanical arm, wherein a front end is connected with a mechanical arm support fixed on a bottom plate, the mechanical arm comprises a flexible pipe, multiple joints and multiple groups of rope yarns, the neighboring joints are spherical surface contact, the flexible pipe passes through all of the joints, the rope yarns pass through all of the joints, and the rope yarns are uniformly distributed along a circumference concentrically installed withthe joints, two rope yarns in one group are arranged at two ends of a diameter of the circumference, an end part of one end of the rope yarn is fixedly connected with the joint at a tail end of the mechanical arm, and the other end passes through the mechanical arm support; a bending mechanism, wherein the bending mechanism comprises multiple rotatable turntables installed on the bottom plate, two rope yarns in one group are fixedly connected with the same turntable after passing through the mechanical arm support, the turntable is rotated so that one rope yarn can be retracted and the otherrope yarn can be released at the same time; and a rigid rod, wherein the rigid rod can be stretched into or stretched out of the flexible pipe, so rigidity of the mechanical arm is changed. The robotis capable of providing rigid support to the mechanical arm, and avoiding a human body from being injured.

The invention discloses a measuring method for the thickness of an extremely-thin covering layer, comprising the following steps of: sampling in a way of being obliquely crossed with a covering layer on a metallographic specimen, and measuring the thickness value of the specimen through a microspore according to Measuring Methods for Thickness of Metal and Oxide Coating by Microscopical Examination specified in the GB6462-86 standard, wherein a manufacturing method of a specimen matrix comprises the following steps of: intercepting a specimen along an intercepting surface vertical to the covering layer to obtain a section as a detected surface, and measuring the thickness value a of the detected surface; and intercepting a detection amplifying surface on a surface obliquely crossed with the surface of the covering layer to form an included angle alpha, and measuring the thickness value b of the detection amplifying surface; and finally, obtaining a measured value delta=bsinalpha.

The invention discloses an interface board, a communication device and a method for configuring interface board. The interface board comprises an access port, an exchange chip, a processor and a relay, wherein the processor is used for controlling the relay to conduct the exchange chip and the access port or conduct the processor and the access port according to the fact that the data transmitting mode is in a bypassed mode, a network bridge mode or a route mode. The technical scheme provided by the invention solves the problems that the existing interface board can not be adaptive to different application contexts and lack of generality and compatibility; according to application requirement, the communication device and the interface board thereof can not only have a data bypassed function, but also have the functions of a network bridge or route data transmission; and on the basis of the reliability of the external interface of the communication device, thus the invention improves the generality and compatibility of the data transmitting functions of the communication device and the interface board.

The invention provides a process for compact natural gas liquefying and floating production and belongs to the field of chemical and cryogenic techniques. In the process, a modular design is adopted, and a liquefaction process consists of propane precooling circulation, double nitrogen expansion cooling circulation, natural gas heavy hydrocarbon removal and liquefaction pipe, wherein propane circulation is used for precooling natural gas and providing a nitrogen coolant; the double nitrogen expansion cooling circulation is used for deeply cooling and liquefying the natural gas; and the heavy hydrocarbon separation is realized by the cold energy generated in liquefaction. The process is suitable for producing liquid natural gas offshore and can prevent the influences of the shaking of a ship on a liquefying process.

The invention discloses an electromagnetic transient implicit reduced order simulation method based on a matrix index. The electromagnetic transient implicit reduced order simulation method comprises the steps of constructing a high-dimensional nonlinear electromagnetic transient simulation model of a power system to be researched under a state analysis framework; setting simulation parameters such as a dimension number m of a dimension reduction sub space, and initializing and starting a simulation program; generating an augmentation state matrix and a state vector within each simulation step length, and generating an orthogonal basis of the dimension reduction Krylov sub space by an Arnoldi algorithm; approximating a high-dimension matrix index of an original system by a low-dimension matrix index according to a Krylov sub space approximating formula of the matrix index, calculating a nonlinear equation to obtain a state variable at the current moment, and boosting one step length for simulation; performing iteration in sequence until simulation is ended. According to the electromagnetic transient implicit reduced order simulation method, high value precision and high rigid processing performance of a matrix index integration method are kept, and general purpose modeling simulation capacity is realized on the nonlinear characteristic of the power system; by the implicit reduced order method, the application range of the matrix index integration method in the field of large-scale power system electromagnetic transient simulation is expanded.

The present invention provides a method for quantitatively evaluating water resource effects of ecological protection and restoration in the river source region. The method comprises the following steps: obtaining soil hydrological characteristic parameters through a prototype ecological hydrological observation test; constructing distributed ecological hydrological simulation of a research area;and finally, according to hydrological characteristic parameters obtained in the preliminary period and the constructed distributed ecological hydrological model, setting scenario plans, comparing theecological hydrological process simulated by each scenario plan with that before the measures are taken, and quantitatively analyzing water resource effects after different ecological protection measures are taken. According to the method provided by the present invention, the problem of applicability of the existing ecological hydrological model in solving the ecological hydrological process simulation in the large-scale, high-cold and high-altitude river source region is solved, and the experimental observation data is directly compared with the model parameters, so that the method can provide a scientific basis for quantitative analysis on water conservation and river runoff effects generated by large-scale ecological protection and restoration measures.

The invention relates to a FLNG / FLPG oil gas pretreatment and liquefaction method which comprises the following steps: 1) arranging an oil gas pretreatment and liquefaction system comprising a natural gas pretreatment system, a light hydrocarbon recovery and separation system, a propane precooling system, a liquefaction supercooling system and a plurality of cold boxes; 2) transmitting natural gas mixture transmitted via a submarine pipeline to a slug flow capturer, then, sending natural gases flowing out of a natural gas pipeline to the natural gas pretreatment system, and sending condensated oil flowing out of a condensated oil pipeline to the light hydrocarbon recovery and separation system; 3) carrying out deacidification, dehydration and demercuration pretreatment on the natural gases flowing out of the natural gas pipeline; 4) carrying out light hydrocarbon recovery and separation on the natural gases obtained after the pretreatment; 5) carrying out propane circulation precooling on the natural gases obtained after the light hydrocarbon recovery; and 6) carrying out liquidation and supercooling circulation processing on the natural gases obtained after the precooling. The method can well solve the problems that in an existing liquidation process, maritime applicability is poor, conventional nitrogen expansion process processing capacity is weak, and efficiency is low. The method can be widely applied to the FLNG / FLPG oil gas pretreatment and liquefaction process.

The invention relates to a cord core drilling tool for a tunnel subhorizontal hole, capable of improving the applicability of the cord core drilling tool in the construction of the tunnel subhorizontal hole. A cord core drilling tool product special for a subhorizontal tunnel is absent in the market at present, only individual user unit for constructing large-angle declined holes improves a ground cord core drilling tool for reluctant use, and the advantages of the cord core drilling technology are not brought into play. By designing a lug boss for fishing a spear and a composite positioning mechanism for opening a clamp and positioning a core shaft, the cord core drilling tool effectively prevents the single-sided clamping in the process of drilling the subhorizontal hole. An inner tube assembly has a sectional structure, the rapid connection and disconnection of an inner tube are realized, and the problem between shorter stroke of the drill and longer inner tube is solved. A mud pump pumps the inner tube assembly and a hydrostatic bailer for providing power for constructing the subhorizontal hole. The cord core drilling tool effectively solves the problem of the applicability ofthe cord core drilling tool during the construction of the subhorizontal hole, and improves the working efficiency of drilling operation.

The invention discloses an automatic workpiece conveying and assembling production system and method suitable for the technical field of automatic production systems. The production system comprises a transmission device, a material accommodating device, a rotating drive device, a first industrial robot and a second industrial robot. The material accommodating device is provided with at least two material storage discs; an image recognition device is arranged above the transmission device and / or the material accommodating device. The production method includes the following steps that photos of different teaching workpieces are taken through the image recognition device, and then the workpieces are alternatively taken out of the material accommodating device and conveyed to the transmission device through the first industry robot; photos of the teaching workpieces on the transmission device are taken through the image recognition device, and then the workpieces are clamped into the material accommodating device from the transmission device through the second industrial robot. The automatic workpiece conveying and assembling production system and method are perfect in adaptability, high in production efficiency and convenient to adopt.

The invention discloses a user interest tag classification method and device, relates to the technical field of the internet, and solves the problem that user information of an open platform cannot be utilized fully and accurately by the third party. The method includes: acquiring training data of active users, wherein the training data include first site data of the active users and user interest tags selected by the active users in a second site; subjecting the second site interest tags to binary tag conversion so as to obtain a binary classifier; creating correlation between the first site data and second site data according to the user interest tags and the binary classifier; acquiring a user interest tag classification model by training through a binary decomposition algorithm and the correlation; computing to obtain the user interest tags of a new user in the second site according to the first site data of the new user and the user interest tag classification model. The method and device is mainly applied to a user interest prediction process of third-party networks or applications.

The invention relates to a device and a method for detecting product pipeline leakage condition. The device comprises a throttling valve (1), a pressure transmitter (2), a connecting assembly (3), a ball valve (4), a tubular main body (5), a pressure transmitter power supply (6), a recorder (7), a safety valve (8) and a blind plate assembly (9), wherein the ball valve (4) is arranged between the pressure transmitter (2) and the tubular main body (5); the pressure transmitter (2) is connected with the pressure transmitter power supply (6) and the recorder (7) to form a loop; the connecting assembly (3) comprises a connecting pipe, a flange interface, a sealing ring, a pipe clamp or a loose joint; the detection device is clamped or screwed on a detected pipeline through the flange interface of the connecting assembly (3); and the recorder (7) is a digital display recorder with alarm function. The device has good sealing property, is accurate and convenient to detect, and overcomes the defects of poor sealing property and sterile pipeline inapplicability of other detection equipment.

The invention discloses a method and a device for generating the information, and relates to the field of knowledge graphs. According to the specific implementation scheme, a plurality of label entitywords are obtained from a target video, the label entity words comprise character entity words, work entity words, video category entity words and video core entity words, and the video core entity words comprise entity words used for representing related content of the target video; aiming at label entity words in the plurality of label entity words, linking the label entity words to nodes in apreset knowledge graph; based on the link result of each label entity word, determining semantic information of the target video; According to the relationship between the nodes and the edges in the knowledge graph, structuring the semantic information of the target video to obtain structured semantic information of the target video. According to the implementation mode, the knowledge graph is adopted to generate the video structured semantic information, and while the video semantic information is reserved, the efficiency and accuracy of videos in recommendation, search and the like can be improved through the structuralization of the semantic information.

The invention relates to the field of marine equipment and discloses an underwater sonar measuring equipment laying and retracting mechanism. The underwater sonar measuring equipment laying and retracting mechanism comprises a base, a laying frame, a laying rod and a rotating shaft, wherein a screw rod bearing seat and a guide rail are fixed to the base; the laying frame comprises a locking hook and an installation frame; the laying rod comprises a first laying connecting rod, a second laying connecting rod and a third laying connecting rod which are connected in sequence; one end of the rotating shaft is provided with a round shaft body, and the other end of the rotating shaft is provided with two hinged semi-arc-shaped hinges; one of the semi-arc-shaped hinges is fixed to the round shaft body, and the other semi-arc-shaped hinge is movable; and after being fastened, the hinges are locked through screws so that the laying rod can be clamped. According to the underwater sonar measuring equipment laying and retracting mechanism, the laying mode is simple, the structure is reliable, after working is completed, equipment can be retracted to a ship side by rotating the laying rod, and the laying and retracting mechanism can sail along with a ship or conduct secondary testing without being disassembled. The underwater sonar measuring equipment laying and retracting mechanism has good suitability and can be carried on the outer sides of ships with various shapes and sizes.

The invention belongs to the technical field of new energy assessment, and particularly relates to a new energy enterprise value assessment model built based on an analytic hierarchy process. The assessment method of the new energy enterprise value assessment model comprises the steps that the analytic hierarchy process is utilized at first, the particularity of new energy enterprises is analyzed, and influences of financial and non-financial driving factors of the enterprises on the new energy enterprise value are quantized to be coefficients; then, the correction coefficients of a stock right free cash flow model are obtained on this basis, the corrected stock right free cash flow model is adopted for calculating the whole value of the new energy enterprises, and the assessment result is more scientific and practical. According to the new energy enterprise value assessment model built based on the analytic hierarchy process, a traditional enterprise value assessment method is inapplicable to value assessment for the new energy enterprises, various driving factors influencing the new energy enterprise value are analyzed in a more comprehensive level, and the assessment result can completely reflect the whole value of the new energy enterprises.

The invention relates to a forecasting method of inner and outer wall temperature difference of a 9% Cr martensitic steel thick wall pipeline in heat treatment. The method comprises the following steps of: calculating to obtain inner and outer wall temperature difference data of T groups of pipelines with different dimensions in postweld heat treatment under the conditions of different heating widths, different temperature preservation widths, and different heat treatment environment temperatures and different temperature control temperatures; establishing a neural network based on an error back propagation by comprehensively considering the influence of the pipeline dimensions, the heating widths, the temperature preservation widths, the heat treatment environment temperatures and the temperature control temperatures on the inner and outer temperature difference in postweld heat treatment and training and testing the neural network; and finally, amending by combining the actual tested data to obtain a method for forecasting the inner and outer wall temperature difference of the 9% Cr novel martensitic heat-resistant steel thick wall pipeline in postweld heat treatment. The forecasting method can rapidly calculate the inner and outer wall temperature difference, can help for instructing and optimizing the heat treatment technology, and can improve the heat treatment quality.

The invention provides a pipeline joint anticorrosion dry film hot melt adhesive. The formula of the hot melt adhesive comprises the following components: by weight, 70-78% of matrix resin, 16-32% of composite tackifying resin, 0.6-1% of a flexibilizer, 0.3-0.8% of an anti-oxidant and 1-3% of a filler, wherein the matrix resin is a mixture of an ethene-acrylic acid ester-maleic anhydride terpolymer (EMH) and an ethene-vinyl acetate copolymer (EVA); the mass ratio of EMH to EVA is 9:1; VA content of the EVA is 28-33%; and melt index (MI) of the EVA is 150-300g / 10min. The invention provides a hot melt adhesive containing matrix resin which is a mixture of EMH and EVA and a preparation method thereof. According to the invention, high temperature resistance of the anticorrosion hot melt adhesive is raised; the applicability problem of the hot melt adhesive under the condition of dry-film construction; and it can be guaranteed that the hot melt adhesive has good bond property under the condition of high temperature.

The invention provides a film surface defect detection method, system and equipment based on deep learning. The film surface defect detection method based on deep learning comprises the steps: S1, conducting image acquisition on the front face of a detected material; S2, carrying out image acquisition on the reverse side of the detected material; S3, performing image preprocessing; S4, carrying out contour appearance defect detection; S5, judging a contour detection result; S6, performing fusion by using multiple channels to obtain a multi-channel fusion image; and S7, importing into a pre-trained deep learning model. The film surface defect detection system based on deep learning comprises an image acquisition module, an image preprocessing module, an image fusion module and an image detection module, thin film surface defect detection equipment based on deep learning comprises a memory and a processor. The method has the beneficial effects that the defects of the optical thin film can be effectively identified on the premise of low cost, the labor cost is reduced, the identification efficiency is improved, the deployment is simple, and the repeatability is higher.