118results about How to "Solving inefficiencies" patented technology

The invention relates to the technical field of medicine quality management, in particular to a medicine quality management method and system based on big data and artificial intelligence, and the method comprises the steps: obtaining the quality influence parameters of a medicine product in the production, transportation and storage processes through a multi-modal data collection device; noise is removed through data cleaning, data dimension reduction is carried out through principal component analysis and an auto-encoder, the analysis efficiency is improved, and the efficiency problem of mass data processing of a traditional method is solved. Time sequence analysis is carried out through the long-short-term memory network, abnormal nodes are recognized in time, manual sampling inspection is replaced, and accurate early warning is achieved. The graph neural network constructs a path tracking model, analyzes the relevance between nodes, traces the quality problem, and reckons the batch of affected drugs. Based on a traceability result, an XGBoost model generates quality risk scores of all links by analyzing historical and real-time data and sets a threshold value, and quality risks are early warned in real time.

The invention relates to the technical field of computers, in particular to the technical field of artificial intelligence, and provides a task processing method and device, electronic equipment and a storage medium which are used for efficiently and accurately performing task planning and execution. The method comprises the following steps: executing at least one round of task processing according to a question input by an object until a preset end condition is met, and outputting reply information corresponding to the question; the nth round of task processing comprises the following steps: planning an nth round of decision action according to a problem and a tool calling result corresponding to an (n-1) th round of target task so as to indicate task planning details of the nth round; extracting the nth round of target task from the task stack according to the nth round of decision action; generating tool calling information for the nth round of target task according to the nth round of target task and the background information corresponding to the problem; and calling a tool according to the tool calling information to execute the nth round of target task to obtain a tool calling result corresponding to the nth round of target task.

The invention provides a cleaning conveying line, which is used for transporting to-be-cleaned parts fixed on a tray and comprises a loading area, a first cleaning conveying area, a second cleaning conveying area and an unloading area. The loading area is arranged at the inlet end of the cleaning conveying line so as to receive and place trays fed by external conveying equipment, and the trays bear to-be-cleaned parts. The first cleaning conveying area is connected with the loading area so as to convey the to-be-cleaned part to a preset cleaning station; the second cleaning and conveying area is arranged on one side of the first cleaning and conveying area so as to receive the parts cleaned by the first cleaning and conveying area; wherein a plurality of translation raceways are arranged between the first cleaning and conveying area and the second cleaning and conveying area, so that a to-be-cleaned part is conveyed between the two conveying areas; and the unloading area is arranged at the tail end of the first cleaning and conveying area, so that the cleaned parts are output to an external production line, and the problem that in the prior art, a cleaning and conveying line is low in efficiency is solved.

The invention provides an event type parallel online learning neuromorphic calculation method and system. The method comprises a framework construction step, a sample learning step, a sparse calculation step and a parallel calculation step. The system comprises a computing core, a Mesh-Tree mixed on-chip routing communication network, a configuration and monitoring module, a global control module and an Argmax decision module. The sparse calculation step (event-driven type) and the parallel calculation step solve the problem of low efficiency when large-scale neuromorphic hardware realizes online learning, and the neural modulation-based online learning framework and the sample learning step have the advantages that the efficiency is high, and the efficiency is high. The compatibility problem and the large-amount memory consumption problem of the current customized online learning processor are relieved; according to the computing system, the computing method is efficiently realized, and meanwhile, the compatibility, the expandability and the learning performance of the neuromorphic processor supporting online learning are improved based on a computing engine architecture capable of configuring a data path and flexible NoC communication.

The invention relates to the technical field of satellite system data transmission, in particular to a satellite system model training method and device, equipment and a medium. The method comprises the following steps: acquiring a distillation data set sent by each distillation satellite; performing centralized training on each distillation data set to obtain a first average gradient; gradient data sent by all the federated satellites are obtained, and a second average gradient of all the federated satellites is generated according to all the gradient data; determining the global gradient of each running satellite; updating the global model parameters according to the global gradient to obtain the global model parameters of the current round, and sending the global model parameters of the current round to the running satellite; and if the global model parameter of the current round does not meet the preset convergence condition, repeating the above steps until the target global model parameter is output to meet the preset convergence condition. According to the method and the device, the problems of overlarge communication load, relatively large privacy security problem and relatively low satellite data analysis capability in the prior art can be solved.

The invention relates to a chip power supply transmission network determination method and device, computer equipment, a readable storage medium and a program product. The method comprises the following steps: acquiring a design training sample of a chip power supply transmission network; the design training sample at least comprises chip layout design information and label data of the sample chip; the label data represents a power supply network template identifier corresponding to each area of the sample chip; according to the design training sample, training to obtain a target chip power supply transmission network determination model; inputting chip layout design information of a to-be-designed chip into the target chip power supply transmission network determination model to obtain a target power supply network template identifier corresponding to each area of the to-be-designed chip; and obtaining chip power supply transmission network information of the to-be-designed chip according to the target power supply network template identifier. By adopting the method, the chip design efficiency can be improved.

The invention relates to the technical field of building steel quality detection, in particular to a steel toughness detection device which comprises a base, two symmetrically-arranged positioning blocks are fixedly installed on the base through a support, downward-pressing sliding grooves are formed in the upper surfaces of the two positioning blocks correspondingly, and the two ends of each downward-pressing sliding groove penetrate through the surfaces of the two sides of the corresponding positioning block correspondingly. A pressing-down sliding groove is formed in the base, a supporting unit is arranged in the pressing-down sliding groove, two symmetrically-arranged fixing rods are fixedly installed on the base, an air cylinder is fixedly installed between the two fixing rods through a support, impact sliding grooves are formed in the sides, close to each other, of the two fixing rods, and an impact unit is arranged between the two fixing rods. The supporting unit and the impact unit are designed, the two ends of the steel pipe can be clamped through the supporting unit, the impact unit is driven by the air cylinder to move downwards, the impact blade impacts the steel pipe with different forces, and the problem that an existing toughness detection machine cannot adapt to the building steel pipe is solved.

The utility model discloses a rail clamping device, a pushing device and a walking system. The rail clamping device solves the technical problems that an existing rail clamping device is insufficient in clamping capacity and inconvenient to position. The rail clamping device comprises a fixed base and a shear fork rail clamping unit hinged to the fixed base, and the fixed base is adjustably connected with a positioning tool used for clamping a rail. The pushing device comprises the rail clamping device; the walking system comprises the pushing device. According to the technical scheme, the rail clamping device is provided with the shear fork rail clamping unit and the positioning tool, before the shear fork rail clamping unit clamps a rail, the whole rail clamping device can be fixed to the rail through the positioning tool, and then the clamping action of the shear fork rail clamping unit can be achieved conveniently; and the positioning tool also has the effect of enhancing the clamping force of the whole rail clamping device. In addition, the positioning tool is adjustable relative to the fixed base, so that the positioning tool can be matched with tracks with different widths, the application range is wide, and disassembly, assembly and displacement are convenient and fast.

The federal knowledge distillation method and application for rubber tire production parameter sharing divide the federal learning process into preheating and distillation two parts, the preheating part adopts the FedAvg federal training strategy, the distillation part utilizes the knowledge generated by each client, carries out the binned formation grouping knowledge, carries out the aggregation after the distribution to each client for the knowledge distillation in the server side.The invention proposes a new federal knowledge distillation method, through the method in the invention, the distributed training of the production line prediction model can be realized, and the model prediction accuracy of the server side is also higher than that of ordinary training.The invention can be used for the distributed training of the rubber tire production line key parameter prediction model, to improve the training efficiency, and in the case of not sharing data, the training result generated by the data can be shared, and the privacy can be well protected.

The invention relates to a fault detection method and device for a graphics processor, equipment, a storage medium and a program product, and the method comprises the steps: obtaining a rasterization file outputted by a rasterization unit in the graphics processor under the condition that a target image outputted by an image output unit in the graphics processor is detected to be abnormal; generating a detection image according to the rasterization file, performing fault detection on the rasterization unit according to the detection image to obtain a first detection result, and detecting other units in the graphics processor according to the first detection result to obtain a target detection result. According to the method, the rasterization file output by the rasterization unit has the initial form of the image, so that the rasterization file can be used for generating the detection image, the detection image is used for carrying out fault detection on the rasterization unit, whether the rasterization unit breaks down or not can be quickly judged, and the fault detection efficiency is improved.

The utility model relates to the technical field of tenoning machines, and discloses a double-end tenoning machine convenient for waste cleaning, which comprises an equipment base, a fixed plate fixedly connected to the surface of the equipment base, the other end of a movable shaft bracket is connected with a movable shaft center, the outer side of the movable shaft center is fixedly connected with one end of a connecting rod, and the other end of the connecting rod is connected with a bearing shaft center; one end of the moving shaft bracket is connected with the shaft rod through the second axis, and the other end of the moving shaft bracket is connected with the connecting rod through the movable axis. According to the utility model, through synchronous operation of the vibration cleaning mechanism and tenoning processing, waste materials are cleaned while processing is carried out, the machine does not need to be shut down for manual cleaning, the production interruption time is greatly reduced, and the overall processing efficiency is improved; and through the linkage design of the cleaning plate and the cleaning and collecting bin, wood chips and crushed materials generated by cutting are pushed into the closed collecting bin in real time, and flying of the wood chips is effectively restrained.

The invention relates to a method for repairing flow data in a user interaction process, and the method comprises the steps: obtaining first data which is basic information of user interaction and comprises a function name of a first processing function; acquiring second data which is a set of to-be-repaired virtual articles; generating third data based on the first data and the second data; finding out an execution entry of the processing function through the function name; traversing the third data, taking the to-be-repaired pipeline data as an input parameter, transmitting the input parameter into an execution entry of the function, and executing the input parameter; wherein the first processing function only processes the received and transmitted flow data of the virtual item based on user interaction, and processing and repairing of the received and transmitted flow data of the virtual item are performed by using the first function. Compared with manual repairing of the user interaction data, the method can effectively improve the efficiency of repairing the user interaction data, and is safe and reliable. The invention further relates to a corresponding device, electronic equipment, a readable storage medium and a computer program product.

The invention relates to the technical field of industrial automation, in particular to automatic line windlass assembling equipment which is characterized by comprising an assembling center, a feeding mechanism arranged on the outer side of the assembling center and used for conveying wheel pieces, and a heating mechanism arranged in the assembling center. The assembly mechanism is arranged on one side of the heating mechanism in the assembly center and used for assembling the row rollers, the heating mechanism comprises a heating device, a rotating disc is rotationally arranged in the heating device, a plurality of clamping assemblies are annularly arrayed on the outer edge of the rotating disc, a fan-shaped disc is arranged at the top of the rotating disc, and the fan-shaped disc is connected with the clamping assemblies. The clamping assembly comprises an installation part, and a sliding rail is arranged at the top of the installation part in a sliding mode. The heating mechanism and the assembling mechanism are integrated, so that continuous work of heating and assembling can be completed, low efficiency caused by a manual assembling mode is avoided, the working efficiency can be greatly improved, and the requirement of large-scale rhythm production is met.

The multifunctional automatic semiconductor equipment system comprises a finished product detection mechanism and a finished product packaging mechanism, and the output end of the finished product detection mechanism is connected with the input end of the finished product packaging mechanism; the finished product detection mechanism comprises a feeding module, a moving module, a detection module, a discharging module and a control module, the moving module is used for moving finished products to the detection module or the discharging module, the detection module is used for testing the finished products, the detection module is electrically connected with the control module, the control module is electrically connected with the discharging module, and the discharging module is used for classifying the finished products. And qualified products are conveyed to a finished product packaging mechanism. The production cost is greatly reduced, the production efficiency is improved, the product detection standard is improved, system collection traceability of test data is achieved, the market failure analysis process is simplified, the product reliability is guaranteed, and the problems that different test links on a traditional production line need different devices to be completed, and the efficiency is low are solved.

The utility model discloses a kind of shaping devices for leisure shoes processing, belong to leisure shoes processing field, including main body rack, the upper end of the main body rack is provided with controller, the upper end of the main body rack is equipped with conveying mechanism, the right end of the conveying mechanism is equipped with conveying drive motor, the upper end of the main body rack is fixedly installed with workstation support, the left side of the workstation support is fixedly installed with rotary drive motor;Through the annular installation design of multiple shoe tree components, then the output shaft of rotary drive motor drives indexer and rotary workstation to rotate, and then drive shoe tree component annular motion, can be operated to multiple pairs of shoes cyclically, then through pressure cylinder driving shaping pressure head plate sliding sequentially to shoe is shaped, greatly improve production efficiency, meet the demand of modernization large-scale production, effectively solve the problem of low efficiency of traditional shaping device.

The utility model discloses an automatic waste discharging device for a stamping die, which comprises a power transmission component, a motion conversion component, a waste discharging component, a waste discharging component, a waste discharging component, a waste discharging component, a waste discharging component, a waste discharging component, a waste discharging component and a waste discharging component. The sliding plate assembly is arranged in the waste falling area of the lower die in a sliding mode and connected with the motion conversion assembly. One end of the elastic reset assembly is fixed to the side wall of the lower die, and the other end of the elastic reset assembly is connected with the sliding plate assembly. According to the utility model, the shaking action of the sliding plate assembly is automatically executed through the power transmission assembly, the motion conversion assembly and the elastic reset assembly, the shutdown for manual intervention is not needed, the real-time discharge of wastes is realized, and the problem of low efficiency caused by manual cleaning in the traditional method is solved.

The utility model discloses a powder surface modification device. The powder surface modification device comprises a reaction kettle, a settling box and a drying box which are connected in sequence, a feeding pipeline is arranged at a position, close to the side wall, of the bottom of the reaction kettle; a first spraying pipe and a second spraying pipe are arranged at the top of the reaction kettle; a stirring motor is arranged in the middle of the top of the reaction kettle, and a first discharging pipeline is arranged in the middle of the bottom of the reaction kettle; the end part of a rotating shaft of the stirring motor is fixedly connected with a dispersing paddle; the second discharging pipeline is arranged above the side wall of the reaction kettle; the second discharging pipeline is connected to the upper part of the settling box; the settling box and the drying box are arranged above the conveying belt, and the drying box is arranged at the position, relative to the rear portion of the conveying belt, of the settling box. According to the continuous powder surface modification device, the continuous powder surface modification process is realized, the problem of low efficiency caused by traditional intermittent production is solved, the production efficiency is improved, and the cost expenditure is reduced.

This technology is applied in the field of network security. This invention provides a method and apparatus for defending against SSH brute-force attacks through log analysis. The purpose of this invention is to solve the problem of inefficiency in existing security defense measures when they cannot cope with brute-force attacks. The main solution includes: acquiring security logs and system logs from monitoring devices and providing them to the perception module; acquiring the attacker's source IP address, username, and time; the perception module recording the attack source IP address, username, and time information in memory and performing statistics within a set time period; querying third-party professional threat intelligence to return the attack source IP address attribution information and threat score; the log threat analysis module defining the danger level of the brute-force threat based on the attack source IP address and threat score; and implementing different corresponding response measures based on the risk identification results of the threat analysis module.

The utility model discloses a seedling lifting shovel, and relates to the technical field of agricultural production. The device comprises a limiting frame, a fixing rod is connected between the inner side walls of the limiting frame, and the surface of the fixing rod is rotationally connected with a fold line plate in a staggered mode through a bearing; the seedling lifting shovel blade is fixedly arranged at one end of the fold line plate, the seedling lifting shovel blade is in a dustpan shape, and the back face of the seedling lifting shovel blade is in an arc shape; the operating rod is fixedly arranged at one end, far away from the seedling lifting shovel blade, of the line folding plate. Due to the adoption of the double-scraper-knife design, soil around seedlings can be reserved as much as possible during digging, falling of the soil is reduced, root systems of the seedlings are protected, and the survival rate after transplanting is facilitated; by adopting the double-handle design, the seedling lifting shovel is more in line with human engineering mechanics, so that the operation is more labor-saving, meanwhile, the accuracy and the stability of the operation are also improved, and the problems of low efficiency, soil falling, root damage and the like of the traditional seedling lifting tool can be effectively solved by using the seedling lifting shovel.

This invention provides a winding device for linear generator coil winding, belonging to the field of generator technology. It includes a base, a left rotary drive system fixed to one end of the base, a right translational pressing system fixed to the other end of the base and moving left and right along the base, and a pneumatic mold system positioned between the left rotary drive system and the right translational pressing system. The pneumatic mold system is connected to the connecting shaft of the left rotary drive system and is directly opposite the pressing part of the right translational pressing system. This invention automates the entire winding process, except for feeding and adjusting the take-up reel, achieving one-time forming. It solves the problem of low efficiency caused by multiple manual "flipping" operations in existing technologies.

The invention discloses an electronic equipment falling protection system based on multi-dimensional sensing. The system comprises a multi-dimensional sensing array, an edge calculation unit, a self-adaptive magnetic field control system and a continuous learning unit, the multi-dimensional sensor array is integrated with a triaxial accelerometer, a gyroscope, a barometer and a magnetometer, and the edge calculation unit adopts a low-power-consumption FPGA and a customized neural network processor to execute a time sequence feature extraction and trajectory prediction algorithm. According to the invention, the problem of insufficient identification precision of a traditional single sensor method under insufficient illumination or complex scenes is overcome; a 5kHz high-frequency sampling rate is adopted, and a multi-stage assembly line parallel processing architecture is combined, so that efficient extraction and trajectory prediction of time sequence characteristics are realized, and the response speed of the system is greatly improved; according to the method, the system power consumption is effectively reduced, and meanwhile, the real-time performance and accuracy of the model are ensured; the multiple electronic devices are controlled to cooperatively generate the dynamic magnetic field network, accurate magnetic suspension track control is provided for the falling device, and the device damage risk is effectively reduced.

The utility model relates to a 40.5 Kv large-current environment-friendly gas-insulated split-box type high-voltage switch cabinet body, which comprises a base, a lower gas box and an upper gas box arranged on the base, a middle radiator arranged between the lower gas box and the upper gas box, a large pressure relief chamber arranged at the front ends of the lower gas box and the upper gas box, a mechanism box arranged at the rear ends of the lower gas box and the upper gas box, and a small pressure relief chamber arranged at the front end of the upper gas box. An instrument box is arranged at the top end of the upper gas box, a lower radiator is arranged on the outer wall of the lower gas box, an upper radiator is arranged on the outer wall of the upper gas box, liquid cooling pipes communicated with each other are arranged in the lower radiator and the upper radiator, a liquid cooling circulator is arranged at the bottom of the lower gas box, and a liquid inlet pipe and a liquid outlet pipe communicated with the liquid cooling circulator are arranged at the bottom of the lower gas box. The liquid cooling radiators are installed on the outer walls of the upper gas tank and the lower gas tank, the liquid cooling radiators are provided with the liquid cooling pipes, the liquid cooling pipes can contain liquid cooling liquid such as water cooling liquid, oil cooling liquid and electronic fluorinated liquid, heat on the radiators is taken away in time through circulation of the liquid cooling liquid, the heat dissipation efficiency is improved, and the problem that a traditional cabinet is low in radiator installation efficiency is solved.

The invention provides a causal analysis method, system and device, electronic equipment and a storage medium. The method comprises the steps that target record information of a to-be-analyzed object and a target causal graph are acquired; the target causal graph is a causal graph obtained by optimizing an initial causal graph generated based on world knowledge through a gradient-based numerical method; the initial causal graph comprises a confusion variable set generated based on world knowledge; and according to the target record information and the target causal graph, generating causal analysis information for the target variable of the to-be-analyzed object. The causal discovery is guided through world knowledge, confusion variables (such as air temperature and seasons) can be automatically identified, and a causal structure is finely corrected by fusing a numerical method, so that the problems that a traditional method depends on manual variable presetting and is low in multi-modal data processing efficiency are solved.

The invention belongs to the technical field of laser processing, and relates to a processing positioning ranging method, which comprises the following steps of: 1) setting a target evaluation area and processing control theoretical parameters; 2) collecting pictures; 3) carrying out definition evaluation on the picture to obtain a definition evaluation result; 4) fitting a definition change curve function based on the definition evaluation result; 5) searching a coordinate corresponding to a theoretical maximum value in the definition change curve function; 6) iterating the coordinates corresponding to the theoretical maximum value to obtain an actual maximum value; and 7) finding out the axis coordinate corresponding to the image with the highest definition according to the actual maximum value, wherein the axis coordinate corresponding to the image with the highest definition is the distance from the camera to the processing position. The invention provides the processing, positioning and distance measuring method which is high in positioning precision, can reduce the use of measuring instruments, reduces the manufacturing cost of processing equipment and can improve the distance measuring efficiency.

The utility model discloses a handheld grinding device for wood processing, which belongs to the technical field of wood grinding and comprises a supporting shell, two groups of telescopic shells, a motor, a transmission gear box, a first transmission mechanism, two groups of second transmission mechanisms and two grinding discs. The two sets of telescopic shells are symmetrically arranged and slidably connected to the two opposite sides of the supporting shell, the motor, the transmission gear box and the first transmission mechanism are all arranged in the supporting shell, and the motor is connected with and drives the transmission gear box; the motor in the supporting shell is connected with and drives the transmission gear box, the transmission gear box is in driving connection with the first transmission mechanism, the first transmission mechanism is in sliding connection with and drives the second transmission mechanism, and the second transmission mechanism is connected with and drives the installation gear shaft. Therefore, the two grinding discs located on the inner sides of the two telescopic shells are driven to rotate and work, and the problem that the efficiency is low due to single-face grinding is solved.

The invention discloses an improved BSIM parameter extraction method based on Dueling DQN, and aims to improve the efficiency and accuracy of parameter extraction in semiconductor device modeling, TCAD simulation data and WAT measurement data are used in the parameter extraction process, parameters related to channel length, width, current and capacitance are extracted, the BSIM parameter extraction problem is modeled as a Markov decision process, and the efficiency and accuracy of parameter extraction in semiconductor device modeling are improved. The method comprises the following steps: establishing a Dueling DQN network model and performing model training, calculating a target Q value and loss, and updating network parameters of the Dueling DQN; in the training process, root-mean-square error values are calculated, invalid actions are removed, even and odd action pairs are processed, sorting is carried out according to the difference of the root-mean-square error values, and converged BSIM parameter values are obtained through model convergence. According to the method, convergence can be carried out in a preset error range, and the efficiency and the accuracy of BSIM parameter extraction are remarkably improved.

The present invention relates to the technical field of hydraulic energy storage starting devices, and in particular to a pressure control method and system applied to hydraulic energy storage starting devices, which also improve the starting efficiency and reliability of the hydraulic energy storage starting device in a complex environment; the method comprises: performing pressure leakage detection on the hydraulic energy storage starting device, and if the pressure leakage rate detected is greater than a preset leakage threshold, a pressure leakage alarm is issued; if the pressure leakage rate detected is not greater than the preset leakage threshold, the starting device state information, the state information of the started device and the real-time environmental monitoring information of the device are obtained; the starting device state information and the started device state information are respectively evaluated in terms of pressure, and a starting device state score for characterizing the operating state of the hydraulic energy storage starting device and a started device state score for characterizing the operating state of the started device are obtained; and operation impact feature extraction is performed on the real-time environmental monitoring information.

This application discloses a composite grinding sander, comprising: a housing; a motor disposed within the housing for providing driving force; a drive shaft connected to and driven by the motor; a connecting seat, one end of the drive shaft disposed within the connecting seat; a oscillating grinding disc disposed at the bottom of the connecting seat, having a grinding disc groove at its center, and connected to the drive shaft via an eccentric transmission mechanism, causing the oscillating grinding disc to oscillate when the drive shaft rotates; and a rotating grinding disc disposed within the grinding disc groove, connected to the drive shaft and capable of rotating with the drive shaft. High-speed rotary coarse grinding is achieved through the central rotating grinding disc, resulting in a high material removal rate; and oscillating fine polishing is achieved through the peripheral oscillating grinding disc, resulting in a smooth surface. Both work collaboratively in the same process, solving both the surface roughness problem of traditional rotary coarse grinding machines and the low efficiency problem of traditional oscillating sanders.

The embodiment of the invention provides a control method, system and platform of an energy storage conversion system. Relates to the power technology field. When the intelligent management module monitors that power generation of the renewable energy system is excessive, the renewable energy system is controlled to charge the supercapacitor bank through the bidirectional direct current-direct current converter, and the supercapacitor bank charges the storage battery pack; when the demand peak of the power grid side or insufficient power generation of the renewable energy system is monitored, the super capacitor bank is controlled to discharge to supply power to the load side, and the storage battery pack is controlled to supply power to the load side after the super capacitor bank finishes discharging; monitoring power data of the energy storage conversion system in real time through a power controller, and controlling power distribution of the energy storage conversion system according to the power data; acquiring working parameters of the super energy storage conversion system, and inputting the working parameters into the intelligent algorithm model to output an optimization strategy; and controlling the energy storage conversion system according to the optimization strategy so as to solve the problem of low efficiency of the energy storage and conversion technology of the power system.

The invention provides a multi-dimensional expansion aviation field large model instruction set generation method, and belongs to the field of instruction set generation, and the method comprises the following steps: S1, generating an instruction set input statement and an output statement based on aviation scientific research literatures; s2, generating a multi-dimensional extended aviation field instruction set; and S3, screening a high-quality aviation field instruction set, and outputting a final instruction set. According to the method, by constructing a high-quality end-to-end instruction data automatic generation and screening method, diversified instructions are effectively generated, the generated instructions are subjected to multi-round and multi-dimensional evaluation, the accuracy and practicability of a final instruction set are ensured, and the method is suitable for popularization and application. The problems of insufficient diversity, difficulty in quality control and the like of automatic generation of the instruction set are solved.