55results about How to "Get rid of constraints" patented technology

The invention discloses a data sharing method, a cloud terminal, a cloud desktop virtual machine and a transit proxy server. When network intercommunication between the cloud terminal and the cloud desktop virtual machine is limited, a terminal side interface module of the cloud desktop virtual machine can establish a first communication link with a client side interface module in the cloud terminal; a virtual machine side interface module can establish a second communication link with a cloud proxy interface module in the cloud desktop virtual machine; a processing module can realize communication interaction between a cloud desktop proxy module of the cloud desktop virtual machine and a cloud desktop client side module of the cloud terminal by utilizing the first and second communicationlinks; therefore, the cloud terminal and the cloud desktop virtual machine are not limited by a desktop transfer protocol, and get rid of the constraint of a network environment; a server device doesnot need to be newly increased in the invention; using habits of users are not changed; data sharing and migration functions of a cloud desktop can be effectively completed; simultaneously, operations related by data sharing are automatically completed by a system; requirements on the IT skills of the users are reduced; and thus, the user experience is improved.

The invention discloses a flaw detection and simulated test device and a flaw detection and simulated test system of a steel rail. The device comprises an emulation computer and a signal generation / data acquisition unit, wherein the emulation computer is connected with the signal generation / data acquisition unit; a steel rail flaw pattern is set by adopting fault feature points to simulate the steel rail with faults; when a simulated test is carried out, a locomotive speed signal is set and ultrasonic echo parameter data is sequentially generated according to the steel rail flaw pattern corresponding to a locomotive driving position and is transmitted to the signal generation / data acquisition unit; the signal generation / data acquisition unit is connected with an outer steel rail flaw detection system and is used for outputting the locomotive speed signal to the outer steel rail flaw detection system, an ultrasonic drive signal sent by the steel rail flaw detection system is received, and an ultrasonic echo simulated signal is sent back to the steel rail flaw detection system according to the ultrasonic echo parameter data transmitted by the emulation computer. With the adoption of the flaw detection and simulated test device and the flaw detection and simulated test system, the technical problems that a test room checking manner for the steel rail flaw detection system does not exist, the testing cost is high and the test is easily restricted by the environment in the prior art are solved; the working performance of the detection system can be checked any time and any where.

The invention provides a cooking method applied to cooking equipment provided with an HSV threshold database. The cooking control method comprises steps as follows: acquiring images of dishes in the cooking equipment at a current cooking stage; converting the images of the dishes into corresponding image signals; extracting food material HSV values contained in the image signals; comparing the food material HSV values with a preset HSV threshold range in the HSV threshold database to recognize food materials contained in the dishes in the cooking equipment; recommending recipes to a user according to the food materials to control the cooking equipment to operate. Correspondingly, the invention further provides a cooking control device, a computer readable storage medium and the cooking equipment. According to the technical scheme, the HSV threshold database is built locally for the cooking equipment, so that the cooking equipment can recognize different kinds of food materials automatically in an environment without a network, and intelligently recommend the recipes to the user according to the recognized food materials, thereby getting rid of network constraint and being simple and convenient to operate, low in cost and high in efficiency.

The invention relates to a surfacing method, in particular to a surfacing method of a 12-percent Cr steel high and middle pressure rotor journal of an ultra supercritical steam turbine, solving the problem of mechanical adhesive wear of the surfaces of the 12-percent Cr steel high and middle pressure rotor journal and a bearing. The surfacing method comprises the following steps of: firstly, respectively turning a first journal and a second journal of a blank rotor in a groove mode; secondly, carrying out dye inspection on a turning surface; thirdly, assembling the blank rotor to a special surfacing device; fourthly, preheating a surface to be surfaced of the rotor journal; fifthly, surfacing the surface to be surfaced; sixthly, distressing and heating a rotor; seventhly, turning and finishing the surfaced rotor; and eighthly, carrying out nondestructive inspection detection on the finished rotor. The surfacing method is used for surfacing the 12-percent Cr steel high and middle pressure rotor of the ultra supercritical steam turbine.

The invention relates to the technical field of floor heating control devices, and particularly relates a distributed electric heating intelligent control system based on the Internet of Things. The distributed electric heating intelligent control system comprises a heating cable, a relay, a digital temperature sensor, a human body infrared sensor, a microcontroller, an independent gateway, a cloud platform and a mobile phone or personal PC equipment, wherein the heating cable is used for supplying heat to a room; the relay is used for controlling the heating cable to operate; the digital temperature sensor is used for monitoring the information of environment temperature of an area; the human body infrared sensor is used for detecting the information whether personnel stay in the area; the microcontroller is used for collecting data information of the digital temperature sensor and the human body infrared sensor and receiving control instructions; the independent gateway is used for sending the collected data information and giving the control instructions; the cloud platform is composed of a cloud server and can be shared by the whole user group; and the mobile phone or personal PC equipment is used for information interaction with the cloud platform and real-time control over an electric heating system. The power cost of power peak time is effectively reduced, energy is automatically stored within power off-peak time at night, and the effects of energy conservation and emission reduction are preferably achieved.

The invention provides a mobile webpay system and a realization method thereof. The system comprises a webpay terminal consisting of transaction equipment and wireless communication equipment, a management website realizing the man-machine interaction between the system and a user and a card transaction front system, wherein the user visits the management website by using a browser of a PC through the internet; the management website is wirelessly connected with the wireless communication equipment; the wireless communication equipment and the transaction equipment are connected with a Bluetooth, serial port or USB in between; the transaction equipment is driven by a client program of the wireless communication equipment; the wireless communication equipment communicates with the card transaction front system through the wireless network; and the card transaction front system communicates with a card transaction background processing system through a DDN dedicated wire. The invention can realize the integrated functions of cash collection, money transfer and capital accumulation through the mobile webpay and has the advantages of quick transaction, wide application and no place limit.

The invention discloses a single-hand operation disabled tricycle which comprises a frame, brake pads, a driving mechanism and two driven rear wheels, wherein the frame comprises a front fork connected with a wheel, a rear fork and a seat; the front fork is rotatablely connected with the frame; the single-hand operation disabled tricycle is characterized in that the driving mechanism is a crank-link mechanism formed by the front fork, a steering lever, a rocker, a crank disc and a connecting rod, wherein the steering lever is fixedly connected with the front fork; the upper end of the rocker is provided with a handle, and the lower end of the rocker is articulated with a vertical surface of the steering lever; the crank disc is connected with the front wheel in a coaxial way; the two ends of the connecting rod are respectively articulated with the peripheral part of the crank disc and the middle part of the rocker; the diameter of the front wheel is larger than those of the rear wheels; and the brake pads are respectively arranged at the two sides of the front fork and the rear fork. The single-hand operation disabled tricycle overcomes the defect that the traditional disabled tricycle can not be driven, steered and braked by a single hand.

The invention provides a cooking control method for setting a cooking device having an HSV threshold database. The cooking control method includes: acquiring a dish image of the cooking device at a current cooking stage; converting the dish image into a corresponding image signal; dividing the image signal into a plurality of matrix regions, and extracting respective food material HSV values included in all the matrix regions; comparing the plurality of food material HSV values and a preset HSV threshold range in the HSV threshold database so as to recognize food materials included in dishes of the cooking device; and recommending a recipe to a user according to the food materials so as to control operation of the cooking device. Correspondingly, the invention also provides a cooking control device, a computer readable storage medium, and a cooking device. The technical scheme establishes the HSV threshold database in the cooking device locally, can allow the cooking device to automatically recognize different types of food materials in a non-network environment, can intelligently recommend a recipe to the user on the basis of the recognized food materials, and is easy and fast tooperate.

The invention provides a taxus chinensis propagating method. The method is characterized by comprising the following steps that 1, taxus chinensis branches are selected; 2, bactericide is used for treating the branches; 3, a plant growth regulator is used for treating the branches; 4, water culture is adopted for inducing the branches to take root; 5, a nutrient solution is added for curing. By the adoption of the method for propagating taxus chinensis, restriction of cultivated land resources can be gotten rid of, and the propagation area is increased; meanwhile, the influence on the survival rate from external environment is reduced to the maximum, time for callus forming and rooting is shortened, the stem cutting survival rate is remarkably increased, the restriction that due to poor environmental factors such as temperature, humidity, soil and water quality, the stem cutting survival rate is influenced by an original propagation means is broken through, the current situation that taxus chinensis genetic resources are scanty can be relieved, and great economic value can be created.

The invention discloses a graph classification method of a cyclic neural network model based on Attention. The Attension idea is applied in the graph classification problem, and the graph classification problem is regarded as a decision process of interaction between a machine and a graph environment in reinforcement learning. Based on the Attention idea, the machine preferentially observes the target area of the graph classification task instead of directly processing the whole graph, so that the target area can be preferentially processed by ignoring the nodes irrelevant to the classification task, and the visual angle movement direction of the machine observation graph can be trained and determined by reinforcement learning rules. At the same time, the model can control the parameters and computational load, and get rid of the constraint on the size of graph data. The invention constructs a cyclic neural network, which integrates the local information of the graph observed before bythe machine through the hidden layer of the circulating neural network, and is used for assisting the decision of the angle of view movement and the graph classification. The invention avoids the problem of subgraph isomorphism in frequent subgraph mining and the problem that the graph kernel function method lacks scalability.

The programmer is connected as: using Forth language processing unit as a core, FPGA connecting directly to Forth language control circuit CPLD, Forth languag processing unit FPGA connecting to logical interface of radio frequency, gradient and external circuit through driving and buffer circuit and Forth language processing unit connecting to static RAM and two-port RAM through driving circuit, ISA interface control circuit CPLD connecting to two-port RAM and PIO and two-port RAM and PIO connecting with bus of ISA. The spectrometer system suitable for NMR, NQR and MRI system can be designed based on the present invention.

The invention discloses a laser ranger indoor calibration device and method based on fiber delay. The device herein comprises a receiving objective, a protective attenuator, a fiber coupler, a variable fiber delay system, a collimating objective, and a variable attenuator unit; the variable fiber delay system comprises a plurality of switches and a plurality of fiber rings different in length; thediameter of a laser beam emitted by a laser ranger under test is decreased through the receiving objective, the laser beam is then partly attenuated by the protective attenuator, the laser beam enters the variable fiber delay system from the fiber coupler and is subjected to delayed transmission, and after light path collimating is performed via the collimating objective and attenuating is performed via the variable attenuator unit, the laser beam returns to a receiving system of the laser ranger under test. The device and method herein are suitable for calibrating maximum ranging distance and ranging precision of a laser ranger.

The invention provides a catering payment integrated management system and an implementation method thereof, and the management system comprises a wireless payment device, a catering management website, a bank card transaction prefix system and embedded industry software, wherein the wireless payment device is used for implementing the functions of magnetic stripe card swiping or IC card read and write, password input, strip code scanning, encryption, packet, wireless communication, and printing transaction receipt, the catering management website is used for implementing man-machine interaction between the system and users, and SCOKET communication method is adopted by the bank card transaction prefix system and embedded industry software. The users use a browser of a computer to access the catering management website through the Internet, the wireless payment device is communicated with the bank card transaction prefix system through wireless, the bank card transaction prefix system is communicated with a card transaction background processing system through a special line, and the bank card transaction prefix system is communicated with the catering management website through a intranet. The invention enables the catering order management and the bank card payment system to combine gapless, and can help an administrator to examine the check-out situation of the users in real time.

The invention relates to an overline operation crane for high-speed railways, belongs to the field of cranes and particularly relates to high-speed railway and rail transit cranes. Two transversing mechanisms including a front transversing mechanism and a back transversing mechanism are respectively erected on four supporting trolleys of a double-line rail, and are in butt joint with the crane through auxiliary supporting legs, and the crane is separated from the original rail line under action of auxiliary supporting leg jacking cylinders, and transverses to the middle of the double-line rail. The front transversing mechanism and the back transversing mechanism moves to be close to the center of the crane together with the front and back corresponding supporting trolleys by retracting effect of the auxiliary supporting legs. Until the auxiliary supporting legs are retracted completely, four crane supporting legs of the crane are unfolded, ball hinges at the tail ends of cylinders of the four crane supporting legs are disposed in ball hinge seats of longitudinal beams of the four supporting trolleys on the double-line rail, and then acting force of the crane to a single-line rail is distributed to the double-line rail successfully, so that operations including overline craning and the like can be carried out for the high-speed railways, and requirements for large-tonnage cranes during rescue and construction on bridge sections of the high-speed railways are met.

The invention discloses an animation generation method. The method comprises the steps of obtaining a target animation model and an animation file; obtaining a coordinate map and a normal map according to the target animation model and the animation file; according to the coordinate map and the normal map, creating an animation material corresponding to the target animation model; when an interface corresponding to the dynamic parameter is called, obtaining a particle parameter set for the target animation model through a particle editor; and generating an animation rendering file set according to the target animation model, the coordinate map, the normal map, the animation material and the particle parameters. The invention also discloses an animation playing method and device, the animation rendering file set can be generated based on the same target animation model, and the animation rendering file set can enable the particle cluster to simulate the motion postures of a plurality ofanimation roles, thereby getting rid of the constraint of a conventional skeletal animation, and improving the animation production efficiency.

The present invention discloses an ESOP minimization method for a logic function. By converting the optimal coverage searching problem of a 3n global space in the ESOP minimization process of an n variable logic function into the simplest connection problem in a plurality of cube blocks, a search space is reduced, thereby breaking away from a constraint of a variable scale; moreover, a cube set is directly operated without being converted into a minterm set, so that limitation to a product term number is avoided; and in order to achieve accurate minimization of rapid ESOP, a minimized conversion algorithm of a cube EXOR conversion diagram is adopted to improve operation efficiency so as to effectively reduce calculation complexity and a memory occupied quantity, and the characteristics that calculation time is insensitive to the number of input variables and is only related to the number of product terms included by the logic function and intersection can effectively achieve an effect that random n variables totally regulate the ESOP minimization of the logic function. The ESOP minimization method has the advantages of no limitation to the number of the product items and the number of the variables in the logic function and capacity of carrying out minimization processing on an ESOP of a random logic function.

The invention provides a switching method in a network virtualization fog wireless access network, being capable of reducing switching signal overhead. The method comprises the steps as follows: a source frog base station judges whether to execute an access switch, if the access switch is executed, the source frog base station sends a user access switch request to a target frog base station; the target frog base station sends command information of allowing user access and wireless resource distribution information to the source frog base station; the source frog base station sends system serial number information to the target frog base station, and forwards cache data to the target frog base station; the frog user is synchronized with the target frog base station, the target frog base station distributes an uplink resource and a tracking area to the frog user, and when the frog user accesses the target frog base station, the target frog base station receives a switch confirmation message from the frog user; and the target frog base station sends a path switching request to a preset virtual sub-group core network, wherein the virtual sub-group core network processes to complete switching. The method provided by the invention is applicable to a radio access network.

Disclosed is a precise advertising-based time aggregation query method. The method comprises the steps that online time data of users is obtained, advertising time period querying requests of advertisers are received, according to advertising time duration of the advertisers and the online time data of the users, effective online time zones are screened out, index structures of all of the time points in the effective online time zones are constructed, and compression is conducted on the constructed index structures. According to the advertising time duration of the advertisers, the online timedata of the users and the index structures of all of the time points are combined, and the best advertising time period which meets the advertising tine duration and allows the most users to be online simultaneously is queried; the function ultimately achieved is that massive online time data of the users on the Internet and mobile Internet medias can be used, and according to the advertising time duration provided by the advertisers, the best advertising time duration is accurately and flexibly determined, so that advertising audiences in the advertising time duration is maximized to improveeconomic benefits of advertising.

The invention relates to the field of robots, in particular to a negative pressure type cascaded underwater wall-climbing robot. The negative pressure type cascaded underwater wall-climbing robot comprises a structural body, a driving assembly, an eddy current sucker, an instrument carrying interface, an underwater control system, mooring ropes, a water surface control system and a detection instrument; the driving assembly, the eddy current sucker, the instrument carrying interface and the underwater control system are arranged on the structural body; the driving assembly, the eddy current sucker and the instrument carrying interface are correspondingly connected with the underwater control system; the structural body, the instrument carrying interface and the underwater control system are connected with the water surface control system through mooring ropes; and the instrument carrying interface is connected with the detection instrument. According to the negative pressure type cascaded underwater wall-climbing robot, the structure is stable; the negative pressure type cascaded underwater wall-climbing robot is safe and reliable; the robot can be stably adsorbed to the surface ofa building body underwater by using eddy current negative pressure and can freely move on the surface of the building body by using the driving assembly; and therefore, stable observation is realized.

The invention provides a draught fan current based air source heat pump defrosting remote monitoring system and a control method, The system involves a current measuring device for measuring a currentsignal of an air source heat pump evaporator fan in real time, a data conversion module used for smoothing, demodulating, decomposing and reconstructing the current signal, acquiring data information, carrying out time domain and frequency domain analysis by using a feature extraction algorithm, and extracting a feature data value, a data diagnosis module used for carrying out signal identification, classification and diagnosis on the feature data value by using a machine learning method, and a defrosting module used for carrying out defrosting control on an air source heat pump evaporator according to the diagnosis frosting condition. According to the draught fan current based air source heat pump defrosting remote monitoring system and the control method, the collected draught fan current data is subjected to data preprocessing and feature extraction, and then data analysis such as the machine learning is combined, so that the outdoor side heat exchanger frosting condition of the air source heat pump is judged, a fault can be found early and cleared, and an air source heat pump unit is guaranteed to operate in an efficient state.

The invention provides a transient high-temperature assembly parameter evolution prediction method for a countersunk bolt mixed connection structure. According to the prediction method, theoretical prediction of the evolution conditions of the pre-tightening force and the nail hole gap of the ceramic matrix composite and high-temperature alloy countersunk bolt connection structure under the high-temperature transient heating condition is achieved through a sequential thermal coupling analysis method, the calculation process is simplified, the structural analysis calculation convergence is improved, important technical guidance is provided for strength analysis and design of the ceramic matrix composite and high-temperature alloy countersunk bolt connection structure, and the method has good practical application potential and can be popularized and applied to multiple technical fields such as aerospace, military and national defense, energy and chemical industry and the like.

The invention provides a guided system of an intelligent tour-inspection device. The guided system of the intelligent tour-inspection device comprises a difference global position system (GPS) device and a control device which controls the running of the intelligent tour-inspection device. The difference GPS device comprises a fixed GPS receiving station and a vehicle-mounted GPS receiving station which is fixed on the intelligent tour-inspection device. The control device comprises a grid setting module which can achieve grid partition for a tour-inspection region of the intelligent tour-inspection device, a movement trend predicting module which predicts the movement trend of the intelligent tour-inspection device, and a route evading module which is connected with a barrier scanning device. Compared with the prior art, the guided system of the intelligent tour-inspection device utilizes the grid technology to set the tour-inspection region, and accurately grasps position information of the tour-inspection device through combining the difference GPS technology. Accordingly, the guided system of the intelligent tour-inspection device gets rid of the restraint of a physical tour-inspection track, and the intelligent degree is improved greatly.

The invention provides a touch event distribution method and device, and the method comprises the steps: obtaining touch events triggered by a current user behavior, and determining the types of the touch events; distributing the touch events to a touch event processor queue; allowing the touch event processor queue to determine touch event processors based on the touch event types so that the touch event processors process the touch events and respond to the touch events on a corresponding user view level, wherein priorities exist among the touch event processors, and each touch event processor is bound with one user view level. In the scheme, the touch events are distributed to the touch event processor queue, so that the touch event processor queue determines the touch event processorsfor processing the touch events according to the types of the touch events, and responds to the touch events on the corresponding user view level. Constraints of view hierarchies can be gotten rid of,the touch events can be accurately distributed to specified view hierarchies according to user intentions, and the use experience of the user is improved.

The invention relates to a turnover box conveying device and belongs to the technical field of light industrial product assembly line production equipment. The device comprises a gutter channel, a movable trolley frame, a movable trolley, a turnover box supporting box lifting mechanism, a turnover box supporting box, a turnover shaft driving mechanism and a turnover box supporting and releasing mechanism, wherein turnover box cavity partition plates are fixed to an upper portion of the gutter channel in the length direction of the gutter channel, the space between every two adjacent turnover box cavity partition plates forms a turnover box cavity, a gutter channel trolley receding cavity is formed between a lower portion of the gutter channel and a terrace, the movable trolley frame is arranged in the gutter channel trolley receding cavity, the moving trolley is arranged on the moving trolley frame, the turnover box supporting box lifting mechanism is arranged on the moving trolley, the turnover box supporting box is connected with the turnover box supporting box lifting mechanism, the turnover shaft driving mechanism is arranged on the front side or the rear side of the gutter channel, and the turnover box supporting and releasing mechanism is arranged between the front side and the rear side of the water flowing groove and connected with the turnover shaft driving mechanism.The device is advantaged in that the requirement that the turnover box rises or falls at the position, corresponding to any workstation on one side of the assembly line groove, of the turnover box according to needs and is transferred to the lower workstation is met, and adaptability to flexible changes of small-batch orders and even special customization is embodied.

The invention discloses a face recognition method and system. The method comprises the following steps that: through perception collection equipment, collecting face image information, labeling the face image information, and obtaining the face image information with label information; sending request service information to a network analysis server; after the network analysis server receives therequest service information, analyzing the request service information to obtain the face image information with the label information; extracting face image feature information from the face image information with the label information, and retrieving face data information from a face database through the label information; comparing the face image feature information with the face data information, and sending a comparison result to terminal equipment; and receiving a comparison result by the terminal equipment, and finishing face recognition on the basis of the comparison result.

The invention discloses a self-flowing asphalt mixture mixer. A mixing pot is arranged on a chassis of the mixer. Three coaxial cylindrical rollers are arranged in the mixing pot and include the outer roller, the middle roller serving as a mixing tank, and the inner roller from outside to inside. A power system and a control system are also arranged on the chassis. Compared with the prior art, the inner roller, the middle roller and the outer roller of the mixing pot are arranged, the middle roller is used as the mixing tank for mixing self-flowing asphalt mixtures, the bottom of the inner roller and the bottom of the outer cylinder communicate to serve as a combustion chamber, the heating area of flames on the surface of the mixing tank is enlarged, and the temperature can be increased more rapidly; and meanwhile, the unique requirement for the mixing temperature (220-260 DEG C) of the self-flowing asphalt mixtures is met. Furthermore, the self-flowing asphalt mixture mixer can achieve the functions of field production and transportation of the asphalt mixtures, gets rid of restraints of a mixing plant, is suitable for low-material-consumption occasions such as curing and repairing of asphalt pavements, and achieves energy saving and high efficiency effects.

The invention relates to a riding assistant with speed recommendation and automatic alarm functions. The assistant comprises a control module, a power supply management module, a positioning module, asensor module, a display module and an alarm module. The control module is electrically connected with the positioning module, the sensor module and the display module. The power supply management module is connected with the positioning module, the sensor module and the display module. The alarm module is electrically connected with the control module. The sensor module comprises a speed measuring sensor module, a temperature sensor module, a heart rate sensor module, an ultraviolet sensor module, a vibration sensor module, an MPU-6050 module and an air pressure module, and the modules are electrically connected with the control module respectively. According to the invention, the speed recommendation and automatic alarm functions of the riding assistant are realized.

The programmer is connected as: using Forth language processing unit as a core, FPGA connecting directly to Forth language control circuit CPLD, Forth languag processing unit FPGA connecting to logical interface of radio frequency, gradient and external circuit through driving and buffer circuit and Forth language processing unit connecting to static RAM and two-port RAM through driving circuit, ISA interface control circuit CPLD connecting to two-port RAM and PIO and two-port RAM and PIO connecting with bus of ISA. The spectrometer system suitable for NMR, NQR and MRI system can be designed based on the present invention.

The invention relates to an Internet of Vehicles based riding aid. The riding aid comprises a control module, a power supply management module, a positioning module, a sensor module, a display module,a keyboard module, a Bluetooth module, a mobile communication terminal, a wireless Internet of Vehicles information transmission module and a cloud server. The control module is electrically connected with the positioning module, the sensor module and the display module, the power supply management module is electrically connected with the positioning module, the sensor module and the display module, and the keyboard module is electrically connected with the control module. The mobile communication terminal acquires various types of information from the riding aid through the Bluetooth module, the wireless Internet of Vehicles information transmission module uploads the various types of riding information acquired by the sensor module to the cloud server, and then data statistics and analysis are carried out. The Internet of Vehicles based riding aid is implemented in order to help riding teams to better know road conditions in riding.

The invention discloses a method for calculating molecular energy of ethane by deep learning, and belongs to the technical field of molecular energy calculation. The method comprises steps as follows:S1, an ethane molecule database is constructed; S2, m configurations and corresponding energy are selected randomly as a training group, and the other (1000-m) configurations and corresponding energyare taken as a testing group; S3, molecular space coordinates of the training group are converted to form a training group configuration parameter input matrix; S4, energy data of the training groupis extracted as a training output energy matrix and the training output energy matrix is in one-to-one correspondence with the training group configuration parameter input matrix; S5, a testing groupconfiguration parameter input matrix and an output energy matrix are constructed, and the line number of the matrixes is 1000-m; S6, the ethane molecule energy matrix E<calc> is obtained by calculation of double nervous layers. According to the method, the influence of a ratio of the training group to the testing group on accuracy of a training result is reduced.