4791results about How to "Achieve protection" patented technology

The invention provides a distributed power data asset management privacy protection encryption architecture design method, which comprises the steps of generating a watermark with robustness and imperceptibility by adopting a content-based digital watermarking algorithm according to statistical characteristics of power data assets, and storing a watermark embedding process and an extraction result through a blockchain; a graph-based data model is adopted, the electric power data assets carrying the watermarks serve as nodes, the data flow relation serves as edges, and an electric power data asset consanguinity graph is constructed; for distributed storage of the blood relationship map, a P2P network topology is constructed by adopting a block chain and a distributed hash table, and blood relationship map data is stored in each node of an alliance chain in a distributed manner; and verifying the digital watermark information embedded in all circulation path nodes through which the power data assets pass in the blood relationship map, calculating the similarity between the extracted watermark and the original watermark, calculating the distortion rate, and judging whether the power data assets are tampered or not.

The invention relates to the technical field of motor control, and discloses a control method and system of a permanent magnet direct current brushless slotless motor. The method comprises the following steps: performing temperature compensation and signal filtering on voltage data of a Hall sensor to obtain rotor position information and temperature characteristic data; performing magnetic field intensity dynamic compensation calculation according to the rotor position and the temperature data to obtain magnetic field intensity parameters; adjusting the amplitude and phase of the three-phase current according to the magnetic field intensity parameter, and generating a current control quantity; extracting and compensating a harmonic component to obtain a torque smooth control instruction; all losses are calculated and optimized, and efficiency optimization parameters are generated; and performing feature extraction and anomaly judgment on the operation parameters, and outputting a protection control strategy. Accurate detection of motor position signals, stable control of torque, optimal control of loss and real-time monitoring and protection of abnormal states are achieved under the high-temperature working condition, and therefore stable operation and the service life of the motor under the high-temperature environment are guaranteed.

The invention provides a privacy protection method and system based on big data security and privacy calculation, and relates to the technical field of data processing, and the method comprises the steps: obtaining original data feature extraction, obtaining data features, selecting a data desensitization algorithm to carry out data desensitization, carrying out encryption processing through a hybrid fully homomorphic encryption algorithm, and carrying out encryption processing through a hybrid fully homomorphic encryption algorithm. Homomorphic encrypted ciphertext data is generated and sent to the privacy protection computing platform; a data preprocessing protocol for privacy protection is constructed, data filtering and data dimension compression are carried out, multi-party security calculation is executed, a ciphertext calculation result is generated and returned to a data provider, verification is carried out, and if verification passes, decryption and filtering are carried out through a decryption algorithm to obtain a pure plaintext calculation analysis result; and constructing a reverse data desensitization model for reduction to obtain an analysis calculation result, constructing a dual-target optimization model for solving, obtaining an optimal balance point and generating an optimal decision result, desensitizing, uploading to a privacy protection block chain network, and performing distributed credible storage and sharing.

The invention relates to the field of power system safety monitoring, and discloses a power transmission line vibration monitoring and early warning method for dealing with extreme weather, which comprises the following steps: arranging a multi-level sensor network to monitor vibration and environmental conditions of a power transmission line; monitoring data are processed and transmitted in real time through edge calculation and a 5G network; performing trend prediction on the vibration data by using a multi-modal deep learning model; an early warning threshold value is adjusted through a distributed self-adaptive dynamic threshold value control method; and when abnormal vibration or a high-risk trend is detected, generating and issuing early warning information. Through a multi-level sensor network, an intelligent early warning algorithm and dynamic threshold control, efficient monitoring of power transmission line vibration and accurate early warning in extreme weather are realized, the real-time performance of data transmission and processing is effectively improved, the accuracy of vibration trend prediction is enhanced, the false alarm rate and the missing report rate are reduced, and the method is suitable for popularization and application. Therefore, the method has higher adaptability and stability in a complex environment.

The invention relates to the technical field of medical image processing, in particular to a medical image enhancement processing method and system based on deep learning, and the method comprises the steps: selecting an image enhancement model according to whether the number of samples needs to be expanded or not; if the number of samples needs to be expanded, performing image enhancement by adopting a generative adversarial network, otherwise, performing image enhancement by adopting an auto-encoder; a federal learning architecture is adopted to train a model, and the model is trained on a plurality of scattered devices or servers. The generative adversarial network is used to generate diversified medical image samples, and the diversity of the samples is improved. And the auto-encoder extracts key features and generates new samples, so that the diversity of data is increased. Through a federal learning architecture, by using dispersed data sources, the diversity of training samples of the model is enhanced, and meanwhile, data privacy is protected. And the diversity of samples is effectively increased, so that the generalization ability of the model is improved, and the method has higher robustness and accuracy in practical application.

The invention discloses a pollution research and water quality prediction method based on river segmented water environment capacity analysis. The method comprises the following steps: accurately dividing a drainage basin into water environment capacity accounting units; obtaining a river water body image through satellite remote sensing; identifying and quantifying the positions and the intensities of point source and non-point source pollution sources in the drainage basin by utilizing the water quality parameters of the river obtained by the remote sensing image; establishing a river water quality prediction model; calculating the water environment capacity of each river reach under different surface water environment quality standards; comparing the estimated pollutant discharge amount with the water environment capacity under each water quality standard, and judging whether the pollutant discharge amount exceeds the capacity or not; and predicting future water quality change by using the model. According to the method, remote sensing, GIS and machine learning technologies are utilized, rapid analysis of the river segmented water environment capacity and accurate prediction of the water quality are achieved, dependence on a large amount of field sampling and complex software simulation is reduced, and the working efficiency is greatly improved.

The invention belongs to the technical field of information recommendation, and discloses a financial information recommendation method and system based on multi-channel distribution, and the method comprises the steps: collecting behavior data of multiple channels of a user, carrying out the cross-platform data integration, and forming a complete user behavior data set; constructing a multi-dimensional user portrait, establishing a financial information database, and marking a keyword tag for each piece of financial information; according to the keyword matching degree of the user portrait and the information content, using content recommendation and financial information conforming to user preferences, and based on historical behavior data of the user and behavior habits of similar users, using collaborative filtering to recommend financial information in which the user is interested to obtain a financial information recommendation list; according to the use habits of the user at different terminals, cross-platform information distribution is performed on the financial information recommendation list through multiple channels, so that the accurate recommendation capability of financial information is improved, and the user experience is improved.

The invention relates to a multi-modal emotion recognition method based on zero-knowledge machine learning, and belongs to the crossing field of artificial intelligence and information security. Key features of local face images, voice signals and posture data multi-modal data are extracted through an attention module, then the key features are transmitted to a server after homomorphic encryption, training of a convolutional neural network (CNN) model is carried out in an encryption domain, and finally a result is sent back to the local for decryption and emotion recognition. The method has the advantages that by using the homomorphic encryption technology, in the whole data processing and transmission process, privacy data of a user is always kept in an encrypted state, and due to fusion of multi-modal features, the emotion recognition accuracy is improved, the robustness of the system in a complex environment is enhanced, the emotion recognition result is more stable and reliable, and the user experience is improved. The method can be widely applied to multiple fields of intelligent monitoring systems, user sentiment analysis, human-computer interaction, mental health monitoring and the like.

The invention discloses rapid die casting equipment for aluminum alloy die castings, and relates to the technical field of metal die casting. A first channel is formed in a die plate and an ejector rod, and a spray hole communicating with the first channel is formed in the ejector rod; according to the device, the mold release agent can be sprayed to the relatively-closed die-casting cavity in a mold closing state, the comprehensive uniformity of the mold release agent during adhesion is improved, the demolding effect after a die casting is formed is improved, the mold release agent is sprayed in the relatively-closed die-casting cavity, and the mold release effect is improved. According to the die-casting mold core, the release agent can be prevented from escaping outwards, the surrounding environment can be protected easily, waste of the release agent can be avoided, energy-saving, environment-friendly, green and efficient die-casting production is achieved, meanwhile, the spraying holes are formed in the cylindrical face of the end of the ejector rod, the spraying holes cannot be exposed when the ejector rod retracts into the die core, and the die-casting mold core is more environment-friendly. And interference to die-casting production due to the fact that the spraying holes are exposed in the die-casting cavity can be avoided.

The invention discloses a security protection method and system for an information management system and a storage medium, and relates to the technical field of security protection.During operation of the system, when a user logs in or carries out sensitive operation, multiple identity verifications are carried out, an identity verification coefficient AC is obtained through calculation, various security parameters in the system are collected and updated in real time, and the security protection efficiency is improved. A high-strength encryption algorithm is applied to data to be transmitted, an encrypted data packet is generated, a transmission encryption coefficient EC is obtained through calculation, corresponding access permissions are set for different data based on user roles, identities and access content, an abnormal monitoring coefficient ADC is obtained, and a comprehensive safety coefficient SEC is obtained through calculation according to an identity verification coefficient AC, the transmission encryption coefficient EC and the abnormal monitoring coefficient ADC. The overall safety of the system is dynamically evaluated, abnormal behaviors are recognized and logs are recorded by comparing the comprehensive safety coefficient SEC with a preset threshold value, and when the system detects an abnormal access mode, a safety alarm is triggered and countermeasures are taken.

The invention discloses a compound enzyme preparation for improving the absorption and utilization rate of a fertilizer and a preparation method thereof, and belongs to the technical field of enzyme preparations. The compound enzyme preparation is prepared from the following components in parts by weight: 100 parts of purified water, 6 to 8 parts of modified cellulase, 10 to 12 parts of immobilized protease, 10 to 12 parts of phosphatase, 10 to 14 parts of urease, 8 to 10 parts of nitrate reductase, 10 to 12 parts of pectinase, 8 to 10 parts of chitinase, 6 to 8 parts of xylanase, 10 to 12 parts of beta-glucosidase, 10 to 12 parts of superoxide dismutase and 4.5 to 6 parts of a microbial agent, the modified cellulase is obtained by chemically modifying cellulase; the immobilized protease is prepared by the following steps: adsorbing activated and coupled modified polylactic acid and a protease solution, crosslinking with glutaraldehyde, filtering, washing and drying. The compound enzyme preparation improves the fertilizer absorption and utilization rate of plants.

The invention discloses a disease prevention management and monitoring system design method based on an adaptive algorithm. The method comprises the following steps: S1, obtaining health data of a user; s2, performing feature extraction and fusion on the health data by using a hypergraph neural network; s3, constructing a disease risk prediction model based on the generative reasoning network, and introducing an anti-fact generation module to generate a health change hypothesis scene; s4, adaptively adjusting a personalized intervention strategy through a nested game model, and adjusting the intervention intensity in the first layer; s5, in the second layer, generating an optimized intervention combination strategy based on a Nash equilibrium driven multi-target adaptive game; s6, risk weighted scoring is carried out on the health data through a structured clustering algorithm, and a feedback event is generated; and S7, dynamically optimizing a disease prevention management strategy through a chained feedback mechanism, and managing node permission. According to the method, the self-adaptive algorithm is combined with the multi-layer game and the distributed anomaly detection, so that personalized disease prevention management and monitoring are realized.

The invention discloses a dry-type transformer. The dry-type transformer comprises a transformer body, a housing, a bottom plate and a fan, wherein the transformer body is formed by a plurality of groups of encapsulated windings in an array mode, a supporting rod is arranged at the bottom of the transformer body and allow the encapsulated windings in each group to connect with each other, the supporting rod is arranged in a crossed mode, and first threaded holes penetrate the supporting rod; the transformer body is integrally arranged in the cavity of the housing, and the bottom of the transformer body is fixedly positioned with the bottom plate, and meanwhile, the transformer body is arranged in the housing through the fan, the whole device achieves dispelling the heat in the transformerbody, and the graphite material of housing side is protruding simultaneously, and the heat from the housing is diffused quickly, and the protrusion of the array distribution can reduce the air great dust entering the inner wall of the housing through air vents.

The invention discloses a non-ideal ISAC system beam forming method with an anti-eavesdropping capability, which is characterized in that a base station is intelligently optimized to transmit beam forming so as to ensure that radar sensing requirements are met under the condition of limited power resources, and attacks of potential eavesdroppers and guards are effectively resisted, so that the anti-eavesdropping capability of the non-ideal ISAC system is improved. Therefore, the safety communication and covert communication performance of system communication users is obviously improved. Specifically, according to the method, a safe ISAC scene which has potential eavesdroppers and guards at the same time, legal nodes of a system have hardware damage, and CSI of illegal nodes cannot be obtained perfectly is considered, and in order to guarantee safe communication and sensing performance, a legal signal energy minimization optimization problem is established. And then, the original problem is converted into a convex optimization problem by utilizing a positive semidefinite relaxation technology and an S process technology, and the convex optimization problem is solved by utilizing a CVX toolbox, so that the strategy effectively ensures the safety communication and target sensing performance of the system under limited power resources, and the complexity of the problem is reduced.

The invention provides a sea surface pollution intelligent monitoring system based on multi-source data, and the system comprises a comprehensive analysis module, an information topology module, a sea surface monitoring module, and the sea surface monitoring module is used for obtaining sea surface pollution monitoring data. The data transmission module realizes interactive communication between the sea surface monitoring module and the information topology module, and the interactive communication comprises transmitting sea surface pollution monitoring data to the comprehensive analysis module; the comprehensive analysis module and the information topological optimization module are used for analyzing and early warning multi-source data, integrating, cleaning and analyzing the data, identifying and correcting errors, abnormities or noise in the data, ensuring the accuracy and credibility of the data, mining data information by using deep learning, then performing intelligent decision distribution, and finally performing data processing. And the monitoring, early warning and emergency response capabilities of sea surface pollution events are improved. Through the above modules, omnibearing, three-dimensional and real-time monitoring and early warning of offshore sudden pollution can be realized.

According to the HBC solar cell wet process provided by the invention, the sequence of a silicon wafer texturing process and a rounding process is optimized, i.e., in the texturing process, the silicon wafer is not subjected to rounding treatment, but is subjected to rounding treatment after etching, so that extra pollution possibly caused by cleaning of a SiNx dielectric film can be removed through rounding treatment; and the texture appearance is ensured. The texturing step is further optimized, a distributed etching process combining texturing etching and polishing alkali etching is adopted to replace a traditional alkali etching process, protection of the front texturing film layer is achieved, relative balance of the front texturing reflectivity and the passivation effect is kept, and the conversion efficiency and manufacturing convenience of the solar cell are improved.

The utility model discloses a battery pack multi-layer protection structure, a battery pack protection plate, a battery pack and electric equipment. The battery pack multi-layer protection structure comprises a bottom protection plate; the tray bottom plate is arranged at one end of the bottom protection plate, and a mounting space is defined between the tray bottom plate and the bottom protection plate; the protection assembly is arranged in the installation space and comprises a buffering piece and an elastic piece, one end of the buffering piece is connected with the tray bottom plate, the other end of the buffering piece is connected with one end of the elastic piece, and the other end of the elastic piece is connected to the bottom protection plate. The elastic piece and the buffer piece are arranged between the bottom protection plate and the tray bottom plate, so that the elastic piece and the buffer piece can be used for buffering and absorbing impact force from the outside in sequence, the battery pack can bear larger impact force, the impact force borne by a battery cell can be reduced, the battery pack is further protected, and the service life of the battery pack is prolonged. And the battery pack can work normally.

The invention belongs to the technical field of joint motors, and particularly relates to a robot joint motor assembly and a motor. The robot joint motor assembly comprises a protective shell and a hollow shaft, a rotor assembly and a stator are arranged in the protective shell, an output flange is fixedly connected to one end of the hollow shaft, a clamping spring is arranged at one end of the hollow shaft, and a fixing plate is fixedly connected to the side, away from the output flange, of the protective shell; and a heat removal assembly and a protection assembly are arranged in the protection shell. According to the robot joint motor assembly, through cooperative use of the rotor assembly, the Z-shaped part and the first rotating wheel, when the robot joint motor assembly is used, the blades can rotate around the central axis of the hollow shaft, airflow flowing in the protective shell is accelerated, heat dissipation of heat in the protective shell is accelerated, and meanwhile in the rotating process of the frame body, the blades can rotate around the central axis of the hollow shaft; the original positions and angles of the blades are correspondingly changed, the guiding effect of the blades on airflow is optimized, and the heat dissipation effect of heat in the protective shell is improved.

The invention provides a network security situation real-time perception and co-processing method, which comprises the following steps of: acquiring a test question downloading log in a campus network, extracting downloading time, user identity, course information and downloading quantity from the downloading log, and discretizing the extracted data to form a scattered batch initial behavior data set; analyzing the early warning signal and the sandbox environment security log, and if the resource occupancy rate suddenly rises, the authority is frequently changed and the network traffic is soaring, judging that the sandbox boundary is maliciously broken through, and obtaining a breakthrough feature analysis result; and dynamically adjusting a sandbox security policy according to the breakthrough feature analysis result, wherein the sandbox security policy comprises limiting the downloading permission of an abnormal anonymous account, isolating soaring malicious traffic, optimizing the allocation of swell resources, and applying the dynamically adjusted sandbox security policy to the sandbox.

The invention discloses a multistage protection technology for an intelligent integrated power supply system, and the technology comprises the steps: monitoring the operation state of the power supply system in real time through various types of configured sensors, and transmitting the monitoring data to a data collection module in a wireless manner for data storage; constructing an internal state analysis model, inputting monitoring data, outputting an internal state analysis value, performing threshold comparison, judging whether the operation state of the power supply system is abnormal or not, and controlling the protection module to be started when the operation state is abnormal; according to the invention, a self-adaptive protection strategy is adopted, the abnormal state of the power supply system can be monitored according to the operation state and environmental factors of the power supply system, various protection measures including short-circuit protection, overload protection, under-voltage protection, overvoltage protection, leakage protection, lightning protection and the like are adopted, comprehensive protection of the power supply system is realized, and the power supply system is protected. The problems that an existing power system protection technology is insufficient in data analysis degree, and protection measure starting conditions are not accurate enough are solved.

The invention discloses a regional adjustable intermittent sampling interference method and system for an unmanned aerial vehicle SAR platform, and the method specifically comprises the steps: generating radar original data according to an intercepted unmanned aerial vehicle carried SAR signal; adopting a Bernoulli mapping sequence to carry out intermittent sampling on SAR signals of the unmanned aerial vehicle, and generating a randomly distributed non-uniform false target group; according to the generated intermittent sampling signal controlled by the Bernoulli mapping sequence, the distance direction and the azimuth direction of the SAR signal are interfered by using a distance direction frequency shift and false motion target characteristic modulation method, and a smart interference signal controlled by the Bernoulli mapping sequence after modulation is generated, so that the interference signal has a false motion characteristic; and after digital-to-analog conversion, low-pass filtering and up-conversion processing are carried out on the generated interference signal, the interference signal is sent by using a transmitting antenna, and interference on the SAR platform radar of the unmanned aerial vehicle is completed. The real moving target in the specific area is protected, and the method has the advantages of being simple, high in real-time performance and good in interference effect.

The invention discloses a silicone oil clutch fan control method and device. The method comprises the steps that an engine air inlet temperature signal, a fan actual rotating speed signal, an engine cooling water temperature signal, an engine torque signal and an engine rotating speed signal are obtained; according to the engine air inlet temperature signal and the fan actual rotating speed signal, the engine air inlet temperature and the fan actual rotating speed are determined; according to the engine torque signal, the engine rotating speed signal and the cooling water temperature signal, the fan duty ratio needed by water outlet temperature cooling is determined, and according to the air inlet temperature signal, the fan duty ratio needed by air temperature cooling is determined; and determining a target fan duty ratio according to the fan duty ratio required for water temperature cooling and the fan duty ratio required for temperature cooling, determining a target rotating speed according to the target fan duty ratio, and controlling the fan to be adjusted to the target rotating speed from the actual rotating speed of the fan. The problems that in the prior art, due to the fact that an electronic control silicone oil clutch fan is imperfect in control strategy, heat of an engine is prone to being accumulated, cooling is not timely, oil consumption is high, and the engine breaks down are solved.

The utility model provides a wheel speed sensor signal analysis circuit, which is used for enabling the circuit to be compatible with a standard current type wheel speed sensor and an AK protocol current type wheel speed sensor at the same time, and comprises an overcurrent protection module, a current sampling resistor and a wheel speed signal output module, the overcurrent protection module is used for closing the connection with the wheel speed signal output module when the signal of the wheel speed sensor is short-circuited; the wheel speed signal output module is used for outputting a level signal corresponding to a wheel speed sensor signal; the overcurrent protection module is connected with one end of the current sampling resistor; one end of the current sampling resistor is connected with the wheel speed signal output module; the other end of the current sampling resistor is grounded.

The invention provides an efficient AI model management and safe loading system and method oriented to a Serverless framework, and aims to solve the high concurrency performance bottleneck and potential safety hazards of AI model loading in a Serverless environment. According to the system, an AI model management module is introduced, efficient semantic cutting is carried out on an uploaded AI model, and fine-grained encryption model blocks are generated; a high-performance P2P network is constructed by utilizing a Serverless cluster, and distributed storage and fast routing of model blocks are realized; a function instance loads a model block in a high-concurrency mode through a P2P network, a multi-stage intelligent cache is constructed locally, and dynamic caching and prefetching are achieved. According to the method, the model loading parallelism degree is improved by adopting a heuristic cutting algorithm, lightweight model updating is realized by utilizing incremental updating transactions, and the loading overhead is minimized. According to the method, the AI model loading performance and efficiency in the Serverless scene are effectively improved, meanwhile, the flexibility and safety of model management are considered, and the method has remarkable technical innovation and practical value.

The invention relates to the technical field of soil sampling equipment, and particularly discloses an automatic whole-section sampling device for saline alkali soil detection, the automatic whole-section sampling device comprises a mounting rack, the inner side surface of the mounting rack is fixedly connected with stabilizer bars, and the stabilizer bars are symmetrically and fixedly connected to the two sides of the mounting rack. The whole-section sampling of soil can be automatically completed by starting the motor, the use convenience is greatly improved, the arranged preparation ring not only can protect the tool bit and prevent the tool bit from being damaged due to collision, but also can provide time for the collection of the sampling plate, and the sampling efficiency is improved. When the motor rotates reversely, the sampling plate spirally moves upwards in the reverse direction, and expands outwards in the upward moving process, so that the soil is released, that is, the soil is released after the device is reset, the operation convenience is further improved, and the working efficiency is improved. The protection effect on the sampled soil is also greatly improved, and the soil is prevented from scattering.

The invention discloses a large model method for network security, which sequentially constructs an adaptive security policy generation model, a network security risk assessment model and a multi-task deep neural network model through multi-stage training, so that a large model covers wide network security tasks, and comprehensive protection from threat detection to risk assessment is realized. The fine training and optimization of each stage ensure that the model deeply understands and deals with complex network security challenges. The self-adaptive security policy generation model adjusts the security policy in real time by learning the change of the network state, and the defense capability is enhanced. The network security risk assessment model generates an accurate risk score by analyzing risk factors, and provides a clear security situation view for managers. A double-attention mechanism is introduced to further optimize feature processing, channel attention intensifies key features, and space attention optimizes space distribution, so that the model is more accurate when capturing and positioning abnormal behaviors.

The invention discloses a preparation method and application of a lactoferrin-sodium alginate hydrogel system, and the preparation method of the lactoferrin-sodium alginate hydrogel system comprises the following steps: S1, preparing a lactoferrin subunit solution; s2, preparing a lactoferrin-active substance co-assembly; s3, purifying; s4, preparing a sodium alginate solution; and S5, preparing the lactoferrin-sodium alginate hydrogel loaded with the active substance. According to the method disclosed by the invention, small-molecule active substances are encapsulated on the basis that lactoferrin has a natural cage-shaped structure and a self-assembly characteristic, and lactoferrin is modified by utilizing good gel characteristic and gastrointestinal digestion resistance characteristic of sodium alginate, so that a lactoferrin-sodium alginate co-protected active substance hydrogel system is constructed; the constructed system shows obvious superiority in processing adaptability and bioaccessibility, and the problems of poor stability, low bioavailability and the like of natural small-molecule active substances in the food processing process are solved.

The invention relates to the technical field of information security, in particular to an information security cloud computing platform based on an intelligent computing center, which comprises a user identity and access control module, a self-adaptive threat detection module, a data protection and backup module, a virtualization and resource management module, an edge computing collaborative security module and a security event management module, the user identity and access control module performs identity verification and management; the self-adaptive threat detection module automatically analyzes and identifies novel security threats; the data protection and backup module encrypts the stored and transmitted data; the virtualization and resource management module manages and monitors the security state of the virtual machine; the edge computing collaborative security module deploys a security agent at an edge node; and the security event management module records security events of the platform. According to the method, the accuracy and real-time performance of threat detection and response are improved, and efficient and safe operation of the platform is ensured.

The invention relates to the technical field of biochar preparation materials, and discloses a preparation method of a modified coal gangue-municipal sludge-based biochar soil conditioner, which comprises the following steps: crushing a coal gangue raw material and a municipal sludge raw material to obtain coal gangue powder and municipal sludge powder; the coal gangue powder and the municipal sludge powder are leached to obtain leached coal gangue powder and leached municipal sludge powder, the leached coal gangue powder and the leached municipal sludge powder are mixed to obtain mixed solid waste powder, the mixed solid waste powder is pyrolyzed to obtain pyrolyzed biochar, the pyrolyzed biochar is acidized to obtain acidized biochar, and the acidized biochar is acidized to obtain acidized biochar. The acidified biochar is subjected to ball milling, and first modified biochar is obtained; active sites of the first modified charcoal are increased, and the target charcoal soil conditioner is obtained. The soil improvement effect of the biochar can be improved.

The invention relates to a privatized large model fine tuning training method and device based on split learning, and the method comprises the steps: splitting a pre-trained large model into a head model segment, a middle model segment, and a tail model segment; splitting the original data into private data and non-private data; according to the privacy data and the non-privacy data, cooperatively training a head model segment and a tail model segment in the client and a first intermediate model segment and a second intermediate model segment in the server to obtain a target large model; wherein in the cooperative training process, a privacy intermediate activation amount which is transmitted between a client and a server and is related to privacy data is subjected to homomorphic encryption; the target large model outputs a document reply about a particular topic. According to the method and the device, the problem that the risk of speculating or recovering the original data through model inversion attack exists due to the fact that the intermediate activation amount needs to be transmitted to the server in the prior art is solved; the privacy data is protected, and the risk of privacy data leakage is reduced.