272 results about "Selection algorithm" patented technology

The invention relates to the technical field of wind power equipment monitoring, and discloses a wind power gear box intelligent fault early warning method and system based on machine learning. The method comprises the steps that multi-source monitoring data such as vibration signals, temperature data and oil analysis data of the wind power gear box are acquired, and multi-scale operation characteristics are extracted through time-frequency conjoint analysis; key fault sensitive features are determined through an adaptive feature selection algorithm, and a dynamic fault feature weight matrix is constructed in combination with a historical fault case library; multi-modal data fusion is adopted to generate an enhanced fault feature set, and modal decomposition is carried out on the enhanced fault feature set to obtain a trend component and a fluctuation component; a fault evolution feature space is constructed by using a deep neural network based on two components, then a fault development mode is identified by using a time sequence mode matching algorithm, and finally a graded early warning signal is generated according to a matching degree with a preset mode, so that fault features can be comprehensively captured, and safe operation of a wind power gear box is ensured.

The invention relates to a priority scheduling generation method and system for Beidou satellite short message communication, and the method comprises the steps: collecting multi-dimensional features, such as message service types, user identities, spatio-temporal information, channel states and the like, dynamically calculating the priority score of each message through a configurable weighting model, and then, carrying out the dynamic calculation of the priority score of each message; the method comprises the following steps of: abstracting a message into a conflict graph vertex, establishing an edge according to a receiving conflict relationship, distributing candidate message sets which are not conflicted with each other for each scheduling time slot by adopting an independent set selection algorithm, and performing joint optimization on a message sending sequence and power distribution by using a genetic algorithm in each time slot by taking a maximized comprehensive utility value as a target, and generating a final scheduling instruction. Through dynamic evaluation, conflict avoidance and resource collaboration, intelligence and self-adaption of a scheduling strategy are realized, and the guarantee capability of a system for key messages and the overall resource utilization efficiency are remarkably improved.

A smart shopping cart identifies items using cameras and sensors. The cart captures images of items within its storage area and applies machine-learning models, such as a barcode detection model, an OCR model, and an image embedding model, to generate identifier predictions. These predictions are processed using an efficient selection algorithm, which may involve majority voting, weighted voting, or linear regression, to select the most accurate identifier. The cart updates its display and user interface with the identified item. The process may be performed primarily by the CPU to enhance computational efficiency, avoiding the latency associated with GPU data transfer. Additional techniques, such as circular buffers and frame skipping, are employed to further optimize resource usage.

The invention belongs to the technical field of software development and testing, and discloses an artificial intelligence-based software development automatic test case generation system, which is characterized in that an immune heuristic case self-repairing module is adopted, defects are regarded as antigens, antibody cases capable of being self-updated are generated by using a clone selection algorithm, and a gene rearrangement mechanism is automatically triggered when an interface is changed, so that the test efficiency is improved. The details of the use case are adjusted while the core detection logic is reserved; compared with a traditional method, the mechanism can realize use case dynamic adaptation without manual intervention, the maintenance workload is remarkably reduced, and the method is particularly suitable for a complex software system with frequent iteration; the space-time coupling test scene generation engine fuses dynamic scenes such as interaction and state transition of a module and short-time operation after precise coverage login by using a space-time convolutional network; the cross-dimension holographic use case synthesis module integrates multi-source data such as codes, hardware and user behaviors through tensor decomposition to generate a composite use case; functions and performance of software in a complex scene can be comprehensively verified, and test blind areas are remarkably reduced.

The invention relates to the technical field of remote sensing data processing and earth surface parameter inversion, and discloses an earth surface matrix parameter hyperspectral data inversion method and system based on machine learning. Comprising the following steps: constructing a multi-source heterogeneous hyperspectral data set; performing feature screening on the preprocessed hyperspectral data set based on an adaptive band selection algorithm, constructing a dynamic weight matrix by calculating mutual information entropy and inter-class distance measurement between spectral bands to realize intelligent screening of key feature bands, and combining spectral derivative conversion and spectral index calculation to generate an enhanced feature vector; and a multi-task transfer learning neural network model is constructed, and an output layer realizes multi-parameter collaborative inversion based on a multi-task learning architecture. And performing preprocessing and feature enhancement operation which is the same as that of the training data on the hyperspectral image data of the to-be-inverted region, inputting the trained neural network model, and outputting a surface matrix parameter inversion result.

The invention provides a distributed photovoltaic data acquisition method and device based on adaptive encryption communication and multi-link redundancy, and the method comprises the steps: collecting real-time power generation data, equipment state data and environment parameters of a photovoltaic system through a multi-source sensor, and forming original data; an MQTT over TLS adaptive encryption communication protocol is adopted to perform end-to-end encryption on original data, and dynamic key management is combined to ensure transmission security; based on a dynamic link selection algorithm, encrypted data is transmitted to a power master station through 4G and LoRaWAN double-link redundancy, so that the transmission reliability is improved; and the master station side decrypts the data and then integrates the data to a database to support power grid dispatching and predictive analysis. According to the invention, the problems of safety and reliability in distributed photovoltaic data acquisition are solved, and the method is suitable for a smart power grid monitoring scene with high reliability and high real-time performance requirements.

A non-contact real-time monitoring system of physiological signs based on millimeter-wave radar includes a millimeter-wave radar and multiple modules for processing radar signals. The millimeter-wave radar is configured to continuously transmit electromagnetic wave signals and simultaneously receive echo signals, perform frequency mixing processing on the echo signals to obtain an intermediate frequency signal, and process the intermediate frequency signal to obtain a radar four-dimensional data matrix. Human body physiological signs are monitored by analyzing body thoracic cavity micro-motion information in signals through the modules; a target echo is processed by adopting a constant false alarm rate detection algorithm, and invalid signals are filtered. A self-adaptive range cell selection algorithm based on short-time stability of respiratory signals is adopted to capture radar echoes reflecting physiological movement. Mixed human body physiological sign signals are processed by using a VMD algorithm, and key parameters in VMD are optimized by using a GWO algorithm.

A method for previewing the terrain in front of an emergency rescue vehicle in a geometric feature degradation scene relates to the technical field of road surface recognition, realizes accurate alignment of LiDAR point cloud and IMU data through timestamp synchronization and linear interpolation, and combines a spherical projection model and a degradation perception complementary feature selection algorithm to realize the previewing of the terrain in front of the emergency rescue vehicle. Converting the three-dimensional point cloud into a robust intensity image and extracting gradient significant features; based on a double-observation secondary filter frame, a motion state is predicted by utilizing IMU forward propagation, a point cloud geometric residual error and an intensity image luminosity residual error are synchronously fused, timestamp deviation is compensated through back propagation, and a multi-source observation model under a global coordinate system is constructed. Finally, the error state is iteratively optimized to realize collaborative output of the high-precision odometer and the three-dimensional terrain map, and the problems of data asynchronism, feature degradation and dynamic interference in a complex scene are effectively solved.

The invention discloses an auction and multi-agent deep reinforcement learning-based electric vehicle charging scheduling method, and relates to the technical field of electric vehicle shared charging reservation scheduling. The method comprises the steps of providing an auction-based shared charging reservation system model; a sharing charging optimal user selection problem is formalized by taking maximization of social welfare as a target; an auction mechanism is introduced, and a new appointment user selection incentive mechanism is provided to calculate a winner set and the payment prices of winner; the system is further expanded, and the reservable time of the charging pile is divided into a plurality of time periods; a charging period selection algorithm based on a multi-agent deep reinforcement learning network is provided, and the objective is to maximize the utility of the charging of the electric vehicle user. According to the invention, the problems of too long waiting time and waste of charging pile resources caused by the fact that the electric vehicle users are gathered in the charging peak period in the existing charging reservation system can be solved, and meanwhile, a multi-period reservation and anti-strategy scheduling mechanism can be provided for the electric vehicle users.

The invention discloses an intelligent autonomous networking monitoring method and system based on optimal channel selection, and a medium, a gateway selects an optimal channel from a plurality of communication channels as a data channel, and divides each time slot of the data channel into a communication segment and a broadcast segment; performing data communication with the networked monitoring node corresponding to the current time slot in the communication section; broadcast information of non-networking monitoring nodes is received in a broadcast segment, and a networking information packet is generated and replied; after being powered on, each monitoring node is switched to a default broadcast channel to send broadcast information to a gateway to request networking, and after receiving a networking information packet, the monitoring nodes which are not networked analyze the networking information packet to obtain networking information, calculate the next wake-up time of the monitoring nodes according to the networking information, and wake up the monitoring nodes when the monitoring nodes sleep to the next wake-up time; the networked monitoring nodes are switched to a data channel to perform data communication with the gateway after being awakened, and the nodes are timed and synchronously awakened with the gateway through a dynamic channel selection algorithm and a time-sharing synchronous communication mechanism, so that the power consumption is reduced.

The invention relates to the technical field of fatigue behavior detection, and provides a GFFY-YOLO-based yawning detection model: on the basis of YOLOv11, an efficient neck global feature fusion network A-GFPN is designed, and a WIoU mechanism is introduced into a loss function part, so that the error of a positioning regression frame is reduced, and the operation speed of the model is improved while high precision is ensured. The method is reasonable and feasible, the key frame reflecting the fatigue state change of the driver can be efficiently and accurately extracted, the detection precision of fatigue behaviors such as yawning and the system response speed are effectively improved, and the calculation burden is greatly reduced while the real-time requirement is met. The method simplifies the video preprocessing process, has good adaptability and expansibility, can flexibly cope with processing challenges brought by different scenes and parameter changes, is suitable for various vehicle-mounted application scenes such as an intelligent cockpit and an ADAS system, has remarkable supplementing and improving effects on an existing fatigue detection technology, and has good application prospects. Wide market prospects and application values are realized.

The invention relates to the technical field of lithium battery SOH estimation methods, in particular to a lithium battery SOH estimation method based on an EIS ensemble learning algorithm. Comprising the following steps: S1, collecting battery performance data; s2, collecting corresponding electrochemical impedance spectroscopy data through an EIS method; s3, obtaining feature data through ICA, DVA and DTV methods, performing normalization processing on the feature data and the data obtained in the S2, and merging the data into a feature vector; s4, calculating a capacity fading rate CAR; according to the capacity fading rate, allocating to different algorithms to carry out SOH estimation, and when the capacity fading rate is less than or equal to 10%, selecting an ELM algorithm to calculate an SOH estimation value; when the capacity fading rate is greater than 10% and less than or equal to 30%, selecting a CNN architecture to calculate an SOH estimated value; when the capacity fading rate is greater than 30%, selecting an SVM algorithm to calculate an SOH estimated value; and S5, displaying and storing the predicted SOC result. Compared with the prior art, the optimal estimation algorithm is dynamically switched based on the capacity fading rate, the full life cycle estimation precision is improved, and the calculation complexity is remarkably reduced.

The invention relates to a polarization reconfigurable array antenna beam forming method using a polarization selection dictionary matrix, and aims to solve the problems that an existing reconfigurable array polarization selection algorithm is low in calculation efficiency, high in parameter sensitivity and incapable of achieving beam scanning. The method comprises the following steps: firstly, giving an expected main polarization direction and a cross polarization direction; then, according to the polarization state which can be provided by each unit, constructing a polarization selection dictionary matrix used for selecting a unit polarization mode; then, according to the polarization selection dictionary matrix and a given polarization direction, constructing a corresponding main polarization beam matrix and a cross polarization beam matrix; and finally, converting a non-convex polarization selection problem into a convexity problem for solving. Compared with an existing stochastic optimization algorithm in the field of polarization reconfigurable array beam forming, the method has higher efficiency and better stability, so that the method has higher practical value.

The invention relates to the technical field of power grid data encryption, in particular to a power grid data hybrid security encryption method based on artificial intelligence driving, and the method comprises the steps: generating a control flow label, a metering flow label and an encryption algorithm selection label according to a power grid data flow characteristic through a scene classification model; wherein the scene classification model is constructed based on a multi-output classification SVM framework; the power grid data traffic features are combined with an encryption algorithm to select labels to be input into a key strength decision model, key recommendation strength and traffic abnormal values are generated, and the key strength decision model is constructed based on a hybrid random forest framework; and based on the control flow label, the metering flow label, the encryption algorithm selection label and the key recommendation intensity, calculating the comprehensive prediction encryption overhead, and setting the self-adaptive hybrid security encryption algorithm of the communication. According to the method, the data security and the encryption overhead are balanced through the SVM-based artificial intelligence driving selection algorithm, and a reliable guarantee is provided for the power grid data security.

The invention provides a soil ammonium nitrogen content hyperspectral prediction method based on an improved extreme learning machine, and relates to the technical field of hyperspectral prediction. The method comprises the following steps: firstly, collecting and treating a soil sample for soil spectral measurement and NH4 < + >-N content determination; measuring soil spectral reflectivity data; preprocessing the soil spectral reflectivity data to form a spectral reflectivity data set; then carrying out characteristic wave band selection by adopting a sequential forward selection algorithm; an improved butterfly optimization algorithm IBOA is adopted to optimize model parameters of an extreme learning machine ELM, and then a hyperspectral prediction model used for predicting the NH4 < + >-N content of the soil is constructed; and finally, the ELM model after parameter optimization is selected to construct a hyperspectral prediction model to predict the NH4 < + >-N content. The method not only provides theoretical and technical support for soil ammonium nitrogen content monitoring, but also provides important reference and guidance for soil nitrogen cycle research and soil management.

A node, in a first network, includes circuitry configured to determine a next hop as decided by Border Gateway Protocol (BGP) is an anycast prefix to a Route Reflector (RR) interconnecting the first network with a second network, responsive to the next hop being the anycast prefix to the RR, create a tunnel with a destination based on the anycast prefix, and utilize the tunnel for traffic having the next hop as the anycast prefix to the RR. The anycast prefix is assigned to two or more RRs interconnecting the first network and the second network. A first path is decided by BGP based on a BGP Path Selection Algorithm that is independent of a second path determined by Interior Gateway Protocol (IGP). The first path and the second path can be different, and wherein tunnel is utilized to ensure the traffic always follows the first path.

The invention discloses an anti-collision method and system based on a multi-band RFID tag, and the method comprises the following steps: calculating an interference distribution characteristic in a target region according to a pre-collected tag parameter; according to the interference distribution characteristics, obtaining an optimal candidate frequency band combination through a frequency band selection algorithm; determining a dynamic adjustment parameter of frequency band allocation according to the optimal candidate frequency band combination; and configuring an RFID tag anti-collision scheme according to the dynamic adjustment parameters. According to the technical scheme, interference analysis, frequency band selection and dynamic adjustment can be automatically processed by a system background, a complete closed loop from original data acquisition to final scheme configuration is formed, and efficient and stable operation of tag frequency band identification in a high-density tag environment is ensured.

The present invention provides a method for selecting cache nodes in an Internet of Vehicles (IoV) under complex scenarios, and belongs to the technical field of IoV. It solves the technical problems of communication interruption and cache node load imbalance caused by obstacles blocking the link and multiple requests from a single vehicle. Its technical solution is: including the following steps: S1: constructing three models; S2: proposing a link quality assessment algorithm; S3: based on the file load upper limit of the cache vehicle, dynamically allocating file requests, and proposing a cache node load balancing algorithm; S4: preferentially selecting cache nodes and task nodes, and proposing a cache node minimization algorithm; S5: giving a cache node selection algorithm process under complex scenarios. The beneficial effects of the present invention are: the present invention considers the maximum number of requests that can be served by the cache vehicle as a load constraint, dynamically allocates file requests, and achieves load balancing.

The invention discloses a rock hardness intelligent identification method based on a vibration signal, and the method comprises the steps: innovatively introducing a genetic algorithm to optimize MPE parameters in a proposed combined noise reduction method based on ICEEMDAN and genetic algorithm optimization MPE, and extracting an IMF which can fully represent the feature information of an original signal; and self-adaptive noise reduction is carried out on vibration signals in the cutting process of the heading machine by combining a wavelet threshold value. According to the Relief-F feature selection algorithm, redundant features can be eliminated and the calculation complexity of the model can be reduced by calculating the correlation weight of the series-connected and fused time-frequency-entropy features and the rock hardness category, and multi-dimensional feature optimization can be realized on the premise of keeping original feature information. A Relief-F method is adopted to carry out dimensionality reduction on time-frequency-entropy fusion features, and original vibration signal feature information is better represented; a CPO algorithm is adopted to optimize an ELM recognition model, the reliability of the model is improved, and compared with an SVM model, a BP neural network model and the ELM model for analysis, it is found that the CPO-ELM recognition model has higher recognition accuracy, and the recognition accuracy reaches 96.25%.

The invention discloses a clear water circulation intelligent scheduling method and system based on water environment situation driving, and the method comprises the steps: carrying out the automatic model configuration based on a multi-mode automatic selection algorithm driven by the water environment situation, and driving the running water scheduling simulation; online simulation calculation of clear water circulation forecasting scheduling and forecasting and early warning of clear water circulation are achieved; a clear water loop scheduling decision-making platform is established by combining a WebGL visualization technology with a WebGis technology, and refined simulation rehearsal of various scheduling scenes on a three-dimensional terrain is realized; through a clear water circulation multi-target scheduling decision scheme optimization method, according to a clear water circulation scheduling result, optimization of a plurality of scheduling schemes is realized, factors such as water change amount, engineering operation time, gate water discharge amount, pump station water diversion amount and the like are brought into an evaluation index system, and a clear water circulation multi-target scheduling decision scheme which not only meets the clear water circulation requirement but also meets the water circulation requirement is optimized. The scheduling time can be saved to the greatest extent, the cost is reduced, and the experience judgment error of a scheduling decision maker is reduced.

The invention relates to a combination model prediction method and system based on time sequence extraction. The method comprises the steps that historical meteorological data and historical wind power data are acquired and preprocessed; constructing a time sequence sample based on the preprocessed data, and extracting time sequence features by using a self-attention mechanism; constructing a diversified model pool containing a plurality of performance complementary basic models, training the basic models, and obtaining a prediction result; performing multi-layer stacking integration on the basic model, and training a stacker model by taking a prediction result of the basic model as input; and performing weighted combination on the stacker model prediction result by adopting an integrated selection algorithm to obtain a wind power prediction result. According to the method, data space-time correlation is fully mined through a self-attention mechanism, multi-layer stacking integration is combined with repeated bagging to improve model generalization robustness, optimal information is aggregated through an integrated selection algorithm, prediction precision is effectively improved, and the requirements of power grid dispatching for wind power prediction precision and reliability are met.

The invention relates to the field of structural health monitoring, in particular to a three-axis sensor number and position optimization method for structural response reconstruction. Aiming at the information coupling characteristics of the three-axis sensor in all directions, constructing a redundancy evaluation index based on the difference degree of a three-axis information matrix, screening the minimum initial measuring point set, removing measuring points with repeated information contribution, and forming a non-redundancy candidate measuring point set; based on a non-redundant candidate measuring point set, a forward greedy selection algorithm is adopted to carry out iterative optimization on the number and positions of sensors, unmeasured redundant measuring points enabling marginal gains to be maximum are selected to be added into an optimization set, it is ensured that redundancy is effectively removed in the initial stage of iteration, the redundancy is moderately relaxed in the later stage, and dependence on prior knowledge of a target structure is reduced; an objective function value changing along with the number of the sensors is used as an uncertainty curve, and the optimal critical value of the number of the sensors is determined by combining an inflection point criterion and a platform criterion.

A system and method for guiding and constraining an artificial intelligence (AI) engine to create and utilize a pre-generated content pool to provide adaptive and personalized learning to users is disclosed. Parsing a user request to identify content requirements and applying an adaptive content selection algorithm that evaluates multiple parameters, including user ID, curriculum standards, content types, and user data. An automated content pool management system maintains a dynamic repository of content aligned with these parameters. Machine learning algorithms enhance and personalize the content pool based on evolving user needs. A large language model (LLM) is employed to generate a guiding prompt that directs the AI engine to retrieve relevant content from the pool. This prompt-driven interaction enables accurate delivery of personalized educational content aligned with user-specific learning goals. The system supports real-time adaptability and individualized content delivery, thereby enhancing the efficacy and responsiveness of AI-powered learning environments.

The invention relates to the technical field of robot cluster control, in particular to a cluster inspection robot task coordination scheduling method based on Beidou short message communication, which comprises the following steps: S1, a scheduling center generates a task list according to an inspection area and task requirements, allocates a task number, a target coordinate and an effective time for each task, and sends the task list to an inspection center; forming a task instruction set and issuing the task instruction set to each robot through a Beidou short message; and S2, after receiving the task instruction set, each robot calibrates a local clock according to the Beidou time service signal, generates a local task queue and executes a corresponding task in an effective time window. According to the Beidou short message forwarding method and system, multi-robot cascade forwarding of Beidou short messages in complex terrains such as mountainous areas, canyons and tunnels is realized by constructing a multi-node relay system and a self-adaptive relay selection algorithm; even if a direct connection signal between a single robot and a satellite is blocked, a relay link can still be established through an adjacent robot to transmit inspection data and task instructions.

The invention discloses a full-process automatic design method and system for a concrete silo structure, and belongs to the technical field of design of concrete silo knots.The full-process automatic design method comprises the steps that S1, a parameter set of the concrete silo structure is obtained; s2, constructing a static calculation engine based on national standard specifications, and automatically executing silo wall internal force calculation, warehouse bottom plate stress analysis and foundation structure mechanical analysis according to the parameter set; the reinforcement area of each part of the silo is calculated; s3, in combination with the calculation result of the reinforcement area of each part of the silo, an intelligent reinforcement selection algorithm is adopted to automatically generate a reinforcement configuration scheme meeting strength, crack and construction constraints; s4, a DXF-format construction drawing and an EXCEL-format steel bar blanking list are automatically generated based on a templated drawing system; and S5, the silo structure model parameters and the steel bar configuration scheme are automatically converted into a BIM model through a Revit secondary development interface, and a complete BIM delivery result containing a concrete member parameterized model and a steel bar three-dimensional model is generated.

This invention discloses a clutter discrimination and suppression method based on multiple echo features, comprising: acquiring radar front-end clutter and target echo detection video data and dividing it into several connected regions; extracting multi-dimensional features and labeling prior information for each connected region; using the Relief feature selection algorithm to calculate and filter the weights of the multi-dimensional features, removing redundant features to obtain the optimal feature vector; using this feature vector to train a support vector machine classifier, and obtaining the optimal parameter model through cross-validation; acquiring measured clutter data, extracting corresponding features and inputting them into the model; and determining whether to remove clutter and retain targets based on the output. This invention effectively filters out dynamic clutter spots, avoids false deletion of weak targets, significantly reduces the false alarm rate of the system, and alleviates the computational burden of subsequent track processing.

This invention relates to the field of information processing technology, and in particular to a gridded interpolation method, apparatus, electronic device, and storage medium that integrates KNN search. It employs a KD-tree with contiguous memory storage and aggregated leaf node storage, which helps reduce memory usage and tree depth. Combined with an improved fast selection algorithm and iterative construction, it can increase tree construction speed and mitigate the risk of recursive stack overflow. By integrating KNN search with heap sort, the nearest neighbor is directly obtained during traversal. Compared to the two-stage method of querying first and then sorting, this reduces intermediate data storage and secondary sorting overhead. This invention improves interpolation speed and reduces memory usage when processing gridded interpolation of millions of scattered points. In typical browser scenarios, it can achieve millisecond-level response times, making it suitable for real-time visualization and rapid analysis in fields such as meteorology, oceanography, and the environment.

The invention discloses a community energy consumption monitoring system based on big data. The community energy consumption monitoring system comprises a data acquisition module, a user typed data processing module, a module for acquiring various types of monitoring sub-models and an intelligent energy consumption monitoring module. The invention relates to the technical field of community energy data processing, in particular to a community energy consumption monitoring system based on big data, which innovatively introduces a community user typed difference analysis mechanism, performs independent learning and differential identification on energy consumption behavior characteristics of different user groups, and improves the energy consumption monitoring precision. A fuzzy lower approximation matrix is constructed based on class variance, and an adaptive feature weight learning mechanism is introduced to construct an energy consumption feature selection algorithm, so that the feature selection precision and stability are improved, and the performance of an energy consumption monitoring sub-model is enhanced; and an optimization algorithm is improved by adopting a three-stage progressive updating strategy and a multi-vector mean value variation strategy, so that the optimization efficiency and precision of the energy consumption monitoring sub-model are improved, and intelligent monitoring of community energy consumption is realized.

A gas temperature field reconstruction method using thermocouple measurement correction is disclosed. First, an Inventor 3D model and an ANSYS temperature field model are established for the gas temperature field. Then, a mapping relationship is established between the control parameters of the gas temperature field and the boundary conditions of the ANSYS temperature field model. Based on the principle of maximizing the influence of fuzzy boundary conditions on the temperature field reconstruction results, experiments are designed for precise boundary conditions and experimental data are obtained. Next, given the boundary conditions of the ANSYS temperature field model, the model is run to obtain the temperature field reconstruction results. A model quality assessment algorithm is run to calculate the fitting degree between the measured curve and the simulation curve, obtaining the algorithm fitting parameters. The algorithm fitting parameters are selected to form the loss function of the gradient descent algorithm. The learning rate, termination condition, and initial iteration parameters are determined. The gradient descent algorithm is run until convergence. Finally, the final iteration parameters are taken as the correction result of the fuzzy boundary conditions of the ANSYS temperature field model. This invention improves the spatiotemporal resolution of the temperature field reconstruction results.

The application discloses a kind of stateful network protocol-oriented fuzz testing method, constructs the corpus of initial test case, and carries out compiling and inserting plug-in operation;Through state selection algorithm, the target state of each round of fuzz testing is selected;From the test case set, the corresponding test case queue is selected, then a test case is selected from the test case queue to generate new test case by variation operation, then the new test case is sent to network protocol server to carry out fuzz testing;After a round of fuzz testing is completed in network protocol server, the feedback information of network protocol server is collected, including code coverage, state coverage and state transition, and subsequent fuzz testing process is guided according to feedback information.The above-mentioned method can solve the problems existing in state selection, test case scheduling and other aspects of traditional network protocol fuzz testing, improve network protocol fuzz testing efficiency and vulnerability mining capability.