47results about How to "Improve availability" patented technology

The application discloses a complex specification multi-modal understanding method and system based on YOLO two-stage segmentation, comprising the following steps: performing pre-processing on an input specification image; inputting the pre-processed image into a coarse-grained detection stage of an improved YOLOv8 network to generate a coarse-grained candidate region set; inputting the coarse-grained candidate region set into a fine-grained segmentation network based on an improved YOLOv8 feature extraction structure; constructing a layout hierarchical relationship structure based on an accurate layout segmentation result; according to the layout hierarchical relationship structure, performing multi-modal feature extraction on the accurate layout segmentation result, and calculating a matching weight; inputting a graphic-text semantic alignment result and the layout hierarchical relationship structure into a logical structure modeling module together to construct a layout element graph; and generating a knowledge graph semantic output result according to a structured logical reasoning result. The application realizes automatic structured understanding of a complex specification by adopting YOLO two-stage segmentation and multi-modal semantic alignment.

The invention relates to a system for realizing transaction gateway distribution based on an asynchronous event model, and the system comprises an SDK access module which carries out the timeout judgment of a transaction request, and receives an asynchronous callback message response; the gateway access module is used for receiving the transaction request and decompressing the data; the service conversion module is used for loading service conversion rules from the service conversion management module and carrying out logic conversion and judgment on the request data; the data storage module is used for storing the business conversion rule and the effective time range thereof; and the service conversion management module provides addition, deletion, modification and check functions of service conversion rules. The system for realizing transaction gateway distribution based on the asynchronous event model has high availability, accesses the server through the SDK access module, realizes background service detection through the heartbeat detection module, realizes automatic switching when a fault is found, and ensures high availability of the service; according to the invention, the method is high in business rule taking-into-effect real-time performance, and has the obvious characteristics of high business rule taking-into-effect real-time performance, high expansibility, high availability and the like.

The utility model belongs to the field of helical gear shafts, particularly relates to a helical gear shaft structure with an adjusting function, and aims to solve the problems that a transmission assembly needs to be completely disassembled during axial adjustment of an existing traditional helical gear shaft structure, axial fixation depends on a single threaded connection structure, and a radial limiting mechanism and an axial fixing device are mutually independent. Comprising a mounting shaft, and the outer wall of the mounting shaft is provided with a plane; the connecting sleeve is arranged outside the mounting shaft in a sleeving manner; the connecting sleeve is sleeved with the bevel gear body in a sliding mode, and a strip-shaped hole is formed in the outer wall of the connecting sleeve. A double-protective-sleeve linkage unlocking structure is adopted, the protective sleeves are rotated to drive the limiting strip-shaped plate to synchronously and transversely move, and the braking state of the pressing plate is rapidly relieved; by means of the three-stage unlocking mechanism of the protective sleeve, the pressing plate and the fixing screws, the steps of a traditional adjusting process are compressed, maintenance personnel do not need to use special tools, and all unlocking operation can be completed only by manually rotating the protective sleeve.

The invention provides a wurtzite boron nitride / diamond composite material and a preparation method and application thereof, the composite material is composed of two phases of wurtzite boron nitride and diamond, and has a stacked sandwich microstructure, the sandwich microstructure comprises at least one first wurtzite boron nitride layer and at least one second wurtzite boron nitride layer, wherein the first wurtzite boron nitride layer is formed by wurtzite boron nitride; at least one second wurtzite boron nitride layer composed of wurtzite boron nitride; the composite material comprises a first wurtzite boron nitride layer, a second wurtzite boron nitride layer and a diamond layer, the diamond layer is arranged between the first wurtzite boron nitride layer and the second wurtzite boron nitride layer and is made of diamond, wurtzite boron nitride is a metastable phase stably existing under normal pressure, and the Vickers hardness of the composite material is larger than 50 GPa. The novel wurtzite boron nitride / diamond composite material disclosed by the invention has high hardness, high toughness and high thermal stability, and the preparation method disclosed by the invention is low in initial raw material price and simple in preparation process.

The invention discloses a preparation method and application of a livestock breeding feed using fruit wine residues as raw materials, which are applied to the field of feed preparation, and aims to solve the problems of low resource utilization rate of fruit wine residues, waste of feed resources, low breeding benefits and the like. According to the method, fruit wine residues are pretreated and precisely compounded, a mineral adsorbent is added to optimize a structure so as to carry out mixing and tempering, multi-stage synergistic fermentation and dynamic regulation and control of a fermentation process, and intermittent stirring and weak forced ventilation are carried out during regulation and control; through the raw material classification and blending, the conditioning technology and the multi-stage synergistic fermentation technology, harmful gas emission in the fermentation process is effectively controlled, the stable fermentation environment is maintained, and the utilization rate of protein and microelements in the feed is remarkably increased; the fruit wine residues which are originally difficult to treat are converted into the high-quality livestock feed with balanced nutrition, so that wine brewing byproducts are turned into wealth, and the problem of environmental pollution caused by wastes is solved.

The invention belongs to the technical field of offshore wind power operation and maintenance and reliability engineering, and discloses an offshore wind turbine generator system-level health degree evaluation method based on dynamic weight distribution, and the method comprises the following steps: S1, collecting and preprocessing the multi-source data of an offshore wind turbine generator; s2, constructing a layered health degree evaluation index system; s3, fusing the mechanism model, the data driving model and the statistical reliability model, and constructing a subsystem degradation model to obtain a subsystem health degree; s4, performing weight fusion in combination with the static weight and the dynamic correction factor to obtain a dynamic weight; s5, fusing the subsystem health degree and the dynamic weight to obtain a system-level health degree; and S6, performing confidence interval sorting and risk grade division according to the system-level health degree. According to the evaluation method provided by the invention, through fusion of a mechanism model, a data driving model and a statistics model, the physical interpretability is reserved, the fitting capability of nonlinearity and coupling degradation is improved, and the misjudgment and missed judgment risks are reduced.

The invention discloses a power station intelligent installation and intelligent maintenance system based on a wireless ad hoc network, and relates to the field of power station operation and maintenance technology and industrial internet technology, and the system comprises a plurality of node devices which are used for data collection, edge calculation, wireless networking and data multi-hop transmission and forwarding; a plurality of node devices form a system through wireless networking, and the system has a distributed multi-copy data storage function and is applied to intelligent installation and intelligent maintenance of a power station. The networking mode of a plurality of node devices in the system adopts a decentralized distributed network, a dynamic source routing protocol and an active bridging protocol are used among the node devices, and the protocols are automatically switched according to the number of the node devices in the network to form a Mesh network. According to the invention, a base station, a wireless router or a switch and other network devices are not needed, and the network can be automatically organized to transmit and share data in the network by taking the network as a collection device and a network communication node device.

The application relates to the field of underwater carrier navigation, and discloses a kind of underwater carrier route and task area conflict detection method.For the problem of high time complexity and low efficiency of route conflict detection algorithm, a dynamic efficient route conflict detection method is proposed.Through pre-processing conflict detection, the possible conflict plan is gradually calculated, a large number of objects without conflict possibility are excluded, the conflict detection processing calculation amount is reduced, and the processing time is saved.The method is suitable for underwater carrier route and task area conflict detection, has high availability and strong practicability.

The invention discloses a multi-level battery system nuclear capacity control method for an electric power frequency modulation energy storage station, and the method aims at the energy storage station which comprises a plurality of transformers, a plurality of energy storage containers hung below each transformer, and a plurality of PCS groups and parallel battery cabinets arranged in each container. The method comprises an energy storage container level capacity checking mode and / or a battery cabinet level capacity checking mode, and capacity checking is achieved by selecting an even number of energy storage containers to carry out internal energy mutual dragging in an energy storage container level; in the battery cabinet layer, an even number of battery cabinets are selected to carry out internal energy mutual dragging so as to realize capacity checking; meanwhile, a one-by-one cut-off charging and discharging control strategy of the direct-current side parallel battery pack is also provided. According to the method, the SOC accumulated deviation caused by long-term frequency modulation operation is effectively solved, the capacity evaluation accuracy is improved, the minimum interference on the frequency modulation service in the capacity checking process is realized, and the system availability rate is remarkably improved.

This invention discloses a heterogeneous hybrid expert large model architecture and alignment method for high-concurrency online medical dialogue. First, the method proposes a dual-pool heterogeneous routing strategy of "4 general experts + 2 departmental experts." Through physically isolated parallel routing channels, it forces the model to simultaneously activate general language capabilities and specialized medical reasoning when processing input, fundamentally ensuring the focus and professionalism of responses. Second, it adopts a structural evolution strategy based on parameter expansion and selective freezing. By completely freezing the pre-trained basic parameters and training only newly added departmental experts, it completely eliminates catastrophic forgetting. Finally, in the model alignment stage, it employs the Group Relative Policy Optimization (GRPO) algorithm for valueless models to reduce memory overhead and designs a dual-track hybrid reward function combining low-level semantic similarity rewards and high-level structured large model referee scoring, effectively solving the "reward hacking" problem and guiding the model to generate refined responses that combine clinical accuracy and humanized interaction. This invention significantly improves the performance of medical dialogue systems. Its sparse activation and parameter freezing characteristics further ensure low latency and high throughput in online high-concurrency scenarios, providing a complete solution for the deployment of reliable medical dialogue systems.

This invention relates to the field of natural language processing and information extraction technology, and discloses a method for automatic generation and iterative optimization of annotation guidelines for event extraction. The method includes: generating initial annotation guidelines based on annotation samples using a large language model and compressing them into lightweight guidelines for predefined event types; employing a two-stage extraction strategy, first identifying event types using the lightweight guidelines, and then recalling complete guidelines based on the types for end-to-end extraction to obtain prediction results; identifying and clustering prediction errors based on a predefined set of event-layer and argument-layer error types to form error clusters; and generating update operations for the annotation guidelines using a large model for each error cluster, and completing iterative optimization after verification of effectiveness. This invention achieves automated generation and continuous evolution of annotation guidelines, and significantly improves the maintainability, generalization ability, and event extraction performance of the guidelines through structured representation and error clustering mechanisms.

The present disclosure relates to the technical field of fan monitoring, and particularly relates to a fan group abnormality monitoring method, system, device and storage medium. The method comprises the following steps: collecting running parameters of each fan in a fan group at a sampling time point; dividing the fan group into at least one reference group according to a division dimension; establishing a group statistical benchmark for the running parameters in a target reference group in the at least one reference group; performing horizontal comparison between the group statistical benchmark and the running parameters, establishing a historical benchmark for each fan, performing vertical comparison between the historical benchmark and the running parameters, and obtaining a comparison result; calculating a relative performance degradation index of a target fan based on the comparison result, and obtaining an abnormality monitoring result corresponding to the target fan when the relative performance degradation index is greater than or equal to a preset alarm threshold. The fan abnormality is detected in advance, the maintenance replaces the emergency repair, and the loss caused by unexpected shutdown is reduced.

The application provides a wurtzite boron nitride / diamond composite material and a preparation method and application thereof. The composite material is composed of two phases of wurtzite boron nitride and diamond, and has a stacked sandwich microstructure, which comprises at least one first wurtzite boron nitride layer composed of wurtzite boron nitride, at least one second wurtzite boron nitride layer composed of wurtzite boron nitride, and a diamond layer composed of diamond and sandwiched between the first wurtzite boron nitride layer and the second wurtzite boron nitride layer. The wurtzite boron nitride is a metastable phase stably existing under normal pressure, and the Vickers hardness of the composite material is greater than 50 GPa. The novel wurtzite boron nitride / diamond composite material has high hardness, high toughness and high thermal stability, and the preparation method has low starting material price and simple preparation process.

The invention discloses an AI-based vehicle scheduling and battery replacement path collaborative optimization method and system, and the method comprises the following steps: collecting multi-source heterogeneous data, including a vehicle state, a traffic road condition and a demand thermodynamic diagram, and recognizing a vehicle deposition point, a shortage point and a battery replacement demand point; uniformly abstracting deposition points, shortage points and battery replacement demand points as space-time network nodes, and constructing a mixed integer programming model with the purposes of minimizing the total operation and maintenance cost and maximizing the operation benefits; inputting a real-time city state into a deep reinforcement learning model, extracting topological features of a road network and task points by using a graph neural network, dynamically calculating the weight of each task node through an attention mechanism, and outputting an optimal task sequence; and converting the optimal task sequence into a work order instruction, issuing the work order instruction to the mobile terminal, and feeding back a monitoring execution result to the credit evaluation system to update the optimization model. The total driving mileage and the labor cost can be reduced, and meanwhile the comprehensive availability rate of the vehicle is improved.

The application provides a method and device for guaranteeing high availability of a mobile terminal application page, and relates to the technical field of mobile terminal application development.The method comprises the following steps: receiving a mobile terminal application page access request sent by a user; loading a bundle resource of a mobile terminal according to the access request; in response to successful loading of the bundle resource of the mobile terminal, showing the mobile terminal application page to the user; in response to failed loading of the bundle resource of the mobile terminal, obtaining a network address of the mobile terminal application page; requesting static resources of the mobile terminal application page from a web server according to the network address and caching the static resources to a local client, and showing the mobile terminal application page to the user.The application automatically degrades to access a corresponding web page when the client environment is abnormal or the bundle resource fails to be requested, avoids the application from being unable to be used due to the client crash, and thus realizes high availability of the client.

The invention discloses a distributed power real-time data access system and method. The system comprises a data ID unified coding module, a unified data access interface module, a fault-tolerant device module and a regional service node module. The data ID unified coding module adopts a structured global unique coding rule, and a code comprises a region code used for directly positioning a data physical storage position; the unified data access interface module receives a client request, groups transmission data, dynamically allocates the request to an optimal regional service agent through a load balancing strategy, and performs data compression, decompression and result aggregation; the fault-tolerant device module realizes overtime processing, failure retry, fault recovery and heartbeat detection functions; the regional service node module is responsible for storage, query and processing; according to the invention, the efficiency and reliability of distributed power real-time data access can be improved.

The invention discloses a parking space state prediction and parking guidance system based on multi-source information fusion, and relates to the technical field of parking guidance, the core of the parking space state prediction and parking guidance system comprises four modules: a multi-dimensional data acquisition and preprocessing module for acquiring and associating three-dimensional data of a user, a parking lot and a road network by deploying a multi-modal terminal and an edge node; the parking demand and resource collaborative evaluation module is used for predicting a parking space vacancy rate by using a model according to the three-dimensional data and judging whether a guiding scheme needs to be adjusted or not by combining a threshold value; the demand-oriented guide parameter optimization module is used for dividing travel scenes based on the core state data of the vehicle owner aiming at a scene needing to be adjusted, calculating an optimal parking path curve and generating a parking space recommendation and path instruction; and the guiding effect closed-loop and compensation module is used for pushing an instruction, monitoring parking duration and evaluating whether a compensation scheme is started or not. According to the system, through multi-source information fusion and dynamic optimization, the parking efficiency and the user experience are improved, and static traffic intelligence is assisted.

The invention provides a multi-communication-source-oriented high-concurrency real-time communication data processing method, and aims to solve the problems of difficulty in compatibility of multi-source heterogeneous protocols, insufficient high concurrency performance, contradiction between real-time performance and orderliness and poor system expansibility in industrial communication. According to the specific technical scheme, a layered architecture of a protocol layer, a routing layer and a logic layer is adopted; protocol adaptation and connection management and control are achieved through an independent protocol adapter, Session unified management and data normalization; decoupling I / O and business processing based on the event-driven asynchronous architecture, the message queue and the working thread pool; by means of accurate routing of a message router, messages of the same kind are bound to the same thread through a consistent hash or modulo grouping strategy, and orderliness is guaranteed. According to the method, high compatibility, easy expansion, excellent concurrency performance, intelligent ordered processing and high-availability elastic expansion are realized, and the efficient data processing requirement of an industrial scene is met.

This invention discloses a SAR target recognition method based on game-theoretic adversarial sample generation. The method involves determining simulated and measured samples, constructing a recurrent generative adversarial network (RGAN) based on an encoder, converter, and decoder, and training the RGAN with the simulated and measured samples to obtain a generator and a discriminator. The simulated samples are then input into the generator and processed to obtain multiple virtual measured samples, forming a virtual measured sample set. The last layer of the discriminator is adjusted, and the virtual measured samples are input into the discriminator to train a target recognition network E. The measured samples are used to fine-tune or correct the target recognition network E to obtain the final target recognition network Q. The target recognition network Q is then used to test validation samples, which are measured data acquired by the radar at a preset or specified time. By utilizing prior target information, this method compensates for the lack of sufficient SAR measured samples, improving SAR target recognition performance under small sample conditions.

The application discloses an intelligent air pump control method, device and equipment and a storage medium, and comprises the following steps: determining whether to directly control the air pump to enter overheating protection according to the actual temperature of the air pump; if not, detecting whether the accumulated working duration of the air pump at a preset temperature meets an overheating condition; when the air pump meets the overheating condition, intervening in the temperature of the air pump in advance to enter the overheating protection to prevent the air pump from directly entering the overheating protection; and when the air pump does not meet the overheating condition, determining whether to stop the air pump from being forced to enter the overheating protection when working for a long time without reaching the preset temperature based on the single working time and the accumulated working duration of the air pump. The application can avoid the high-heat condition of reaching the highest temperature, increase the serviceable rate of the air pump, reduce the waiting time and improve the experience of continuously using the air spring system.

A kind of charging device for sharing massage device storage device, including outer casing, the storage box for carrying waistband type massager is retractably arranged in outer casing, the power of battery is guaranteed by external socket, the bottom of storage box is fixedly provided with support plate, support plate is formed with the load groove for placing waistband type massager at the center of storage box, the charging port for charging wire to go out is opened in the bottom center of load groove;Between battery and charging port, charging wire is provided, charging wire includes charging head in charging port and soft wire for connecting charging head and battery.The beneficial effects of the utility model are: charging port is set in the bottom center of load groove, optimization returns process, user does not need additional adjustment position to aim at charging head, improves convenience;Soft wire connects charging head and battery, can flexibly bend or stretch when storage box retracts and moves.

The application relates to the technical field of computer software, and discloses a heterogeneous asset dynamic discovery and mixed loading rendering method and system of a low-code platform. A three-dimensional logical addressing coordinate system is constructed by using module identification, business entity identification and page identification; in the construction stage, a code directory is scanned, multi-dimensional identifiers corresponding to physical assets are identified according to file levels and naming rules, and a logical mapping registry is generated; in the running stage, a rendering scheduling center queries the registry based on the logical addressing coordinate system, and performs cascading detection between native code assets, local static JSON descriptions and remote dynamic API interfaces; when the native code assets are hit, native code components are synchronously called as UI skeletons, and dynamic metadata of business entities are asynchronously pulled in parallel; and the dynamic metadata is bound to the native code components through a context injection mechanism. The application takes into account high-performance rendering of native code and dynamic real-time updating of business logic.

The application relates to an artificial intelligence technology and discloses a data import template configuration method, which comprises the following steps: constructing an import demand matrix according to data import demands; calculating the matching degree between the import demand matrix and a plurality of preset import rules, and selecting an import rule with a matching degree greater than a preset matching threshold as a rule to be screened; obtaining business data corresponding to the data import demands, and obtaining data import conditions corresponding to the business types; constructing a rule screening model according to the data import conditions, and screening a target import rule from the rule to be screened by using the rule screening model; obtaining a blank data import template, and rendering the data import template by using the target import rule to obtain a standard data import template. In addition, the application also relates to a blockchain technology, and the data import demands can be stored in nodes of the blockchain. The application further provides a data import template configuration device, an electronic device and a storage medium. The application can improve the usability of the generated data import template.

The embodiment of the invention relates to an inference service instance binding method and system, and the method comprises the steps: obtaining instance identification information of at least one candidate inference service instance through an inference client in an inference session establishment process; under the condition that the historical instance identification information is locally stored in the reasoning client, matching the locally stored historical instance identification information with the instance identification information of at least one candidate reasoning service instance; if the matching is successful, determining the successfully matched candidate inference service instance as a target inference service instance, and establishing or maintaining session connection with the target service instance to carry out an inference dialogue; wherein the historical instance identification information is instance identification information of inference service instances connected to the inference client in the last round of inference session; according to the method, when the inference session is newly built, the client can be directly connected to the original instance in which the historical cache is stored, computing resources are reused, meanwhile, a centralized component does not need to be introduced, and the performance bottleneck and the single-point fault risk are avoided.

This application relates to the field of nuclear power technology, and in particular to a nuclear power plant start-up and shutdown method, its control system, electronic equipment, and storage medium. The nuclear power plant start-up and shutdown method according to this application requires first obtaining the operating sequence procedure and fault handling procedure; in response to receiving a nuclear power plant start-up and shutdown command, sequential control operations are executed on each controlled nuclear power device according to the sequential control steps in the operating sequence procedure; in response to detecting abnormal information about the nuclear power plant during the execution of the sequential control operation, the sequential control operation is stopped, the current sequential control node is recorded, and an anomaly feedback operation matching the abnormal information about the nuclear power plant is executed according to the fault handling procedure; in response to detecting that the abnormal information about the nuclear power plant has disappeared, the sequential control operation continues from the current sequential control node; in response to all sequential control steps in the operating sequence procedure being executed, the target nuclear power plant is confirmed to be in a start-up or shutdown state, thereby improving the availability of unit operation while reducing the risk of misoperation.

The invention discloses a standard compound library MetaTag database and application thereof.43120 compounds are recorded in the MetaTag database, the compounds have Xm-Yn type chemical structural characteristics, X represents an electrophilic substituent, Y represents a nucleophilic substituent, m and n represent the number of X and the number of Y respectively, the character '-' represents that X and Y form a covalent bond, and X and Y are both from endogenous metabolites; each compound is composed of a virtual part and an entity part; in the virtual part, each compound corresponds to a unique entry, and each entry comprises but is not limited to sub-entries such as a chemical information class, a biomedical information class, a related literature or an external database link; in the entity part, each compound can find a molecular entity in a mixed standard series of a MetaTag database, and the molecular entity is used for collecting LC-MS data to confirm the metabolite structure.

The angle steel drilling production line based on power head walking comprises a fixing mechanism and a drilling mechanism, the fixing mechanism comprises first electric sliding rails which are symmetrically arranged, electric sliding arms are installed on the first electric sliding rails in a sliding mode, U-shaped seats are rotatably installed at the ends of the electric sliding arms, the U-shaped seats are connected with a material fixing seat, and the material fixing seat is connected with the drilling mechanism. A right-angle supporting plate is arranged on the material fixing seat, a side clamping plate and an upper clamping plate are slidably mounted on the outer side of the right-angle supporting plate, and the drilling mechanism comprises second electric sliding rails which are symmetrically arranged; the right-angle supporting plate is matched with the inner angle of the angle steel and is matched with the pneumatic clamping plate, so that the angle steel drilling machine can quickly adapt to the angle steel with different lengths and different placement angles, the universality is high, three-axis high-precision drilling is realized, the drilling machine head is driven by an X-axis, Y-axis and Z-axis electric sliding rail system, the positioning precision is high, the moving speed is high, and the working efficiency is high. And continuous and efficient machining of complex hole position patterns can be achieved.

This utility model discloses a lower end bearing device for a trough-type scrubbing machine, including a lower end bearing device comprising a lower end shaft, a shaft sleeve, an adapter, a bearing housing, a self-aligning roller bearing, and a double-hole shaft end retaining ring. This enables the device to achieve an ultra-long sealing life: employing a labyrinth seal (zigzag gap) to block mud and sand particles, combined with two lip seals to prevent slurry penetration, significantly improving the physical barrier capability against high-concentration mud and sand, preventing solid impurities from entering the bearing (extending the sealing life by more than 2 times compared to traditional structures, reaching the standard of 12 months of continuous operation); enhanced bearing operational stability: replacing sliding bearings with self-aligning rolling bearings eliminates the risk of dry friction caused by sliding friction, reducing frictional resistance by 40%-60% and reducing operating energy consumption by 15%-20%; and a double reduction in maintenance costs and downtime losses: reducing the annual maintenance frequency from 4-6 times to 1-2 times, increasing equipment availability by 20%-30%, and significantly reducing labor and spare parts replacement costs.

The invention relates to an AI-based automatic control method and device for a thermal forging ingot machining process, and the method comprises the steps: obtaining a cross-section image and temperature information of a forging ingot through a thermal imaging dual-light camera, transmitting the cross-section image and temperature information to an AI detection server to obtain defect information, and executing defect treatment measures when defects are detected; and when no defect is detected, edge pixel coordinate data is obtained, the actual size deviation is calculated, error parts and types are determined, and forging operation is executed after forging process parameters are obtained. The device comprises a manipulator, a camera, an AI detection server, an application server, a client front end and a forging press which are accessed to a service network through a switch. By means of the AI-based automatic control method and device for the hot forging ingot machining process, the problems that in hot forging ingot measurement operation, the labor condition is severe, the measurement time is long, the measurement result is poor, data is limited, the treatment cost of forging ingot defects is high, the forging and pressing treatment cost of forging ingots is high, and the feeding treatment cost of the forging ingots is high can be solved.