120results about How to "Reduce computing pressure" patented technology

The invention provides a concrete production management method and system based on digital twinning, and relates to the technical field of digital twinning, and the method comprises the steps: collecting the health monitoring data of concrete production equipment in real time through a plurality of sensors, and collecting the full life cycle event data of the concrete production equipment; preprocessing the health monitoring data at an edge end, extracting state features of the equipment, performing real-time diagnosis by using a pre-trained fault prediction model based on the state features of the equipment, and constructing an equipment multi-level digital twinborn model at a cloud platform; the method comprises the following steps: fusing a pre-constructed physical degradation model and a data driving model based on a multi-level digital twinborn model, constructing an equipment residual life prediction model, and dynamically adjusting a concrete production strategy in the digital twinborn model according to an obtained prediction result to obtain an optimal concrete production scheduling scheme. And the information is converted into an equipment control instruction, so that automatic issuing and execution of a concrete production scheduling scheme are realized.

The invention discloses a coal mine carbon emission prediction method and system based on digital twinning and LSTM, and the method comprises the steps: building a carbon emission source information coding standard based on a business demand and a digital twinning model of a mining area in combination with a production link and an auxiliary production link of the mining area; performing carbon emission edge calculation on the carbon emission energy consumption data of the carbon emission source by using a coal mine carbon emission accounting method to obtain carbon emission data; establishing a corresponding mapping relation between the coal mine digital twinborn model and a database of carbon edge computing equipment by utilizing a carbon emission source information coding standard; according to the mapping relation, mapping the carbon emission data of each link to a coal mine digital twinborn model one by one; and establishing a carbon emission prediction model by using an LSTM algorithm, inputting historical actual carbon emission data into the model, and predicting carbon emission data of the current day. According to the method, the current carbon emission situation of each link of coal mine production and auxiliary production can be known, the carbon emission trend of each emission point is mastered in real time, and the emission reduction strength of enterprise energy-saving measures is quantified.

The invention discloses a virtual power plant adjustable potential comprehensive evaluation method, system and equipment giving consideration to communication and action time delay. According to the method, a communication system time delay model, an equipment action time model and a communication reliability model are constructed by constructing a cloud edge collaborative network system framework for regulating and controlling a virtual power plant. And then, establishing a virtual power plant maximum adjustable potential evaluation model based on timing delay step length search by using a k-shortest path algorithm. And finally, providing a method based on a near-end strategy optimization algorithm to solve the optimization model. According to the method, comprehensive evaluation of the adjustable potential of the virtual power plant considering the influence of the communication system and the resource action time is accurately described, a decision basis is provided for the virtual power plant to participate in primary and secondary frequency modulation and peak regulation of the power system, and frequency stability and power and electric quantity balance of the power system can be supported.

The invention discloses a power plant electrical automation control system and method based on hierarchical management, and relates to the technical field of power plant electrical control, and the system comprises a dynamic resource distribution coordination mechanism, a fault self-healing mechanism, and a local optimization control and energy efficiency optimization strategy based on edge calculation. The resource output of the dynamic resource allocation coordination mechanism is used as the optimized resource input of the local optimization control based on the edge computing, and the energy efficiency optimization mechanism is used as a dynamic interaction layer of the dynamic resource allocation coordination mechanism and the local optimization control based on the edge computing. The fault self-healing mechanism is used as a result execution mechanism of the energy efficiency optimization strategy. By designing a dynamic resource allocation coordination mechanism, the resource load dynamic allocation function of the power plant equipment is realized, the problems of unbalanced resource utilization, equipment overload or low-efficiency operation and incapability of quickly responding to resource requirements in a traditional power plant are solved, resource allocation is optimized, and the operation efficiency of the whole system is improved.

The invention belongs to the technical field of data analysis, and particularly relates to an information analysis method for bidding documents. The information analysis method comprises the following steps: S1, obtaining a newly added bidding and tendering file, and sending the file into a preprocessing module; s2, the preprocessing module outputs corresponding preprocessing data to a classification module; s3, the classification module divides the content in the current preprocessed data into valid tables, core blocks and independent blocks, and then the valid tables, the core blocks and the independent blocks are sent into a table analysis module, a core analysis module and an independent analysis module respectively; and S4, the table analysis module, the core analysis module and the independent analysis module respectively output the first analysis data, the second analysis data and the third analysis data to an integration module for integration to obtain core information and output the core information. According to the method, the core information of various bidding and tendering files can be accurately and efficiently analyzed.

The invention discloses an obstacle-aware cloud-side-end cooperative industrial wireless network channel prediction method, and relates to the technical field of wireless communication. The industrial terminal equipment acquires channel state data in real time, preprocesses the channel state data by using a data preprocessing unit, and transmits the preprocessed channel state data to the edge server; based on the preprocessed channel state data, the edge server adopts a long-short-term memory network to carry out obstacle detection and classification, generates obstacle category information, and uploads the obstacle category information and the preprocessed channel state data to a cloud server; and an obstacle sensing channel conversion architecture model is established, and the cloud server obtains a channel state prediction result through the obstacle sensing channel conversion architecture model based on the received obstacle category information and the preprocessed channel state data. The method is high in accuracy, higher in adaptability in a complex environment, and good in reliability and expansibility in practical application.

The invention provides a data query method and device, electronic equipment and a storage medium, and relates to the technical field of databases, the method comprises the following steps: in response to an obtained query instruction, determining a query tag corresponding to the query instruction; determining a corresponding target field in a plurality of local field sets based on the query tag; the target field is a related field of a target data set; wherein the plurality of field sets are attribute information sets of the data set periodically updated from the server; fields representing the same service attribute in the multiple field sets correspond to the same label; and sending the target field to the server, and receiving a target data set fed back by the server in response to the target field. According to the method, the efficiency of querying data from the server by the terminal is improved, and the accuracy of querying the data is improved.

The invention relates to the technical field of picture rendering, in particular to a media picture rendering optimization method and system based on artificial intelligence. The method comprises the following steps: acquiring a terminal video stream; performing RGB format conversion on the terminal video stream to generate an RGB type frame image set; pixel partitioning is carried out on the RGB type frame image set, and RGB pixel matrix blocks are generated; performing color space acceleration conversion on the RGB pixel matrix block to generate a YUV cache data matrix; the YUV storage data matrix is sent to a DSP internal processing unit for key frame selection, and YUV key video frame data and YUV non-key video frame data are obtained; and performing intra-frame coding on the YUV key video frame data so as to generate YUV key frame coded data. According to the method, the media picture rendering efficiency is improved by optimizing data processing, transmission and rendering and combining user visual focus analysis.

The invention provides a remote sensing image slicing system and method based on an FPGA (Field Programmable Gate Array). The system comprises a CPU (Central Processing Unit) module and an FPGA module, the CPU module is used for decoding an input original image and transmitting decoded image data and a corresponding first configuration parameter to the FPGA module; the FPGA module comprises an image slicing module and a memory module; the image slicing module is used for segmenting the original image into a plurality of slice images based on the first configuration parameter, and the memory module is used for storing the original image and the generated slice images. A CPU module and an FPGA module are combined, a plurality of image slices can be processed at the same time by means of the parallel computing capacity of an FPGA, and the processing speed and efficiency are improved; and the method has relatively low processing delay and flexibility.

The invention relates to the technical field of battery car charging management, in particular to a battery car charging integrated management system based on edge calculation, which acquires battery parameters of a battery car and state parameters of a charging pile in real time through a battery car end acquisition module and a charging pile end acquisition module, and performs local processing through an edge calculation unit. The real-time performance of data processing is improved, and abnormal conditions can be quickly responded. Charging behaviors are analyzed through deep learning, abnormal conditions are judged, and data are distributed to a load balancing unit or a risk assessment unit. The load balancing unit utilizes a reinforcement learning model to optimize charging load distribution according to the power grid load, the battery car demand and the charging pile state so as to avoid power grid overload. The risk assessment unit establishes a model through historical data, monitors the risk in real time and gives an alarm, and meanwhile, through edge calculation and local deployment of a classification model, the dependence on the cloud is reduced, and the calculation efficiency and the management efficiency are improved.

The invention discloses a large language model training method based on efficient fine tuning of parameters. The large-scale language model training method comprises the following steps: aiming at a large-scale language model of a training object, dividing the large-scale language model into a client pre-training model and a server pre-training model; the client pre-training model and the server pre-training model are combined for parallel cooperative training, and in the training process, the client pre-training model and the server pre-training model are subjected to parallel training by adopting an efficient parameter fine tuning technology; after training is completed, combining the client pre-training model and the server pre-training model by adopting a model aggregation method; and repeating the steps until the large language model serving as the training object is converged. The large language model training method is based on the segmentation federal learning framework, and combines the parallel training advantage of the FL and the model segmentation advantage of the SL, so that the training efficiency is greatly improved.

The invention particularly relates to a GIS mass data drawing method and system. According to the GIS mass data drawing method, a ServiceWorker thread is started in the background, a GIS data interface is intercepted through the ServiceWorker thread, and GIS total data is intercepted; selecting part of data according to a user-defined standard, pushing rendering data to the main thread, and rendering the rendering data as initial data; dividing the remaining total data into a plurality of data sets according to a user-defined preset rule, and putting the data sets into a queue to be rendered; pushing the to-be-rendered data to a main thread in batches by utilizing an incremental updating mode; and after the main thread receives the data, calling a rendering method to render the data to a page. According to the GIS mass data drawing method and system, the calculation pressure of a main thread can be relieved, long-time blank waiting and long-time occupation of a rendering thread can be avoided in an incremental rendering mode, and therefore instant huge pressure brought by direct rendering of mass data to equipment is avoided.

The invention discloses an internet-based online gynecological disease intelligent consultation system, and relates to the field of data processing, and the system comprises the steps: generating a public and private key pair of a patient through employing an RSA asymmetric encryption method; a BERT model based on deep learning is adopted to process the decrypted inquiry information, processing is carried out on local equipment, and structured patient condition description is obtained; the encrypted patient condition description is uploaded to a server, a system private key is used for decryption on the server, diagnosis is carried out by using a reasoning method based on a knowledge graph according to the patient condition description, and a disease candidate list is obtained; transmitting the encrypted candidate list to local equipment of the patient, decrypting by using a private key of the patient, and reasoning by adopting a CART decision tree according to the disease candidate list to obtain a final diagnosis result; aiming at the low precision of online gynecological disease consultation diagnosis in the prior art, asymmetric encryption, knowledge graph reasoning, a CART decision tree and the like are adopted, so that the diagnosis precision is improved while the privacy of a patient is protected.

The invention discloses a positioning data intelligent management system and method applied to an Internet of Things product, and relates to the technical field of positioning data management, and the system comprises a distribution area analysis module, a distribution task distribution module, and a distribution unit management module. The distribution area analysis module is used for acquiring a distribution area image and map data, dividing a distribution area into path nodes and path units, marking the path nodes and the path units, and constructing a distribution path node graph; the distribution task distribution module analyzes the distribution exchange rate of the shortest distribution path of each distribution task, carries out the least grouping decision of each distribution task, and plans the distribution path of the distribution unit of each group. And the distribution unit management module determines the path units where the distribution units are located according to the real-time positioning data of the distribution units, and sets obstacle avoidance strategies when the distribution units advance according to the calculation results of the environment change rates of the path units.

The invention relates to the technical field of transportation monitoring, in particular to a transportation monitoring system based on artificial intelligence video analysis. Comprising a data acquisition unit, an end data storage unit, a monitoring platform, a video preprocessing unit, a video analysis unit, an early warning unit and a display unit. The video preprocessing unit is used for judging whether the condition is normal or not, and the video analysis unit performs enhancement processing on video data and judges whether a fault exists or not and the fault type; the early warning unit is used for analyzing the probability of future fault occurrence under the fault-free condition; according to the invention, the collected video data is processed step by step, and the video data is enhanced, so that the features of the video data are more obvious, and the training of a preset model and the output of a result are facilitated; according to the invention, an artificial intelligence algorithm is adopted to analyze the collected video data and output a conclusion, so that artificial interference is reduced; and the output conclusion can be directly embodied on the video and is more visual.

The invention discloses an original data change processing method for a power transaction data analysis platform, and belongs to the technical field of new energy station power generation. Complete original data and a blood relationship link are formed, key business fields are reserved in all data structures, and when the updating time of the original data is changed, the original data is updated. Real-time or off-line scheduling can trigger a blood relationship task, and all downstream nodes of the whole blood relationship link can modify data of all tree nodes according to the change of updating time; according to the method, a solution and a thought are provided for a data change scene of an electric power spot transaction decision-making auxiliary system, the difference between a traditional counting warehouse and an electric power transaction industry counting warehouse is analyzed, and the problem is analyzed and the solution is given through the perspective of a standardized model, data synchronization and the like.

The invention provides an intrusion detection system and method for an edge device level lightweight pruning neural network, the scheme provides a lightweight pruning strategy of a neural network, the neural network can be easily deployed on edge devices while the detection performance of the neural network is ensured by reducing the scale of a neural network model, and the detection efficiency of the neural network is improved. The computing pressure of the edge device is effectively reduced, so that the edge terminal device can autonomously detect the abnormal traffic, and the problems that the detection efficiency is low and the detection is easily influenced by a network communication environment because the detection can be carried out only by a cloud in a traditional scheme are effectively solved.

The application provides a distributed urban drainage monitoring method, system, device and medium, relates to the technical field of urban drainage monitoring, and comprises the following steps: acquiring monitoring data of a plurality of monitoring nodes in a target drainage area; according to a topological connection relationship of an urban drainage pipe network, the monitoring nodes are divided into at least one local topological subnet; the number of local topological subnets is configured with a corresponding edge computing node; the edge computing node determines a first monitoring result based on the monitoring data of the monitoring nodes and corresponding standard monitoring data; if the first monitoring result is abnormal, the self-state of the monitoring node is pre-checked, if it is determined that the monitoring node has a self-fault, the monitoring data of the monitoring node is reported to a central cloud platform based on the edge computing node; if it is determined that the monitoring node has no self-fault, the monitoring node is marked as a suspected event node; a corresponding edge computing node triggers a distributed collaborative diagnosis process to obtain a second monitoring result and report the second monitoring result to the central cloud platform.

The invention discloses an electric vehicle distributed charging optimization method and system based on a carbon emission sensitive dynamic electricity price, and the method comprises the steps: setting an initial iteration strategy based on a controllable electric vehicle load model, and averagely distributing the needed charging amount to all charging time periods; defining the size of an algorithm convergence criterion according to the required convergence precision and the number of the electric vehicles; constructing a dynamic electricity price containing coupling of a load item and a carbon emission item, and calculating an EVn cost function according to the dynamic electricity price; starting a kth round of iteration, traversing all EVn in each round of iteration, and counting k from 0; other electric vehicle charging power vectors except the EVn are calculated; after each iteration, checking whether a convergence criterion is met or not; if yes, the obtained load curve is an approximate solution of Nash equilibrium, and if not, the charging strategy is updated repeatedly; and after the load curve under the dynamic electricity price mechanism is obtained, the load curve under the time-sharing static electricity price with the same electricity price level is obtained through calculation. The method is suitable for electric vehicle group charging management in a community scene.

The present invention relates to artificial intelligence technology, and discloses a data skew correction method, including: receiving a data skew correction instruction input by a user, extracting an original data set from multiple servers according to the data skew correction instruction, performing data amplification processing and adding random number processing on the original data set to obtain a data set to be partitioned, performing a partition operation on the data set to be partitioned to obtain a partitioned data set, performing statistics on the partitioned data set based on a pre-built tuple evaluation model to obtain a tuple partitioned data set, performing data skew correction on the tuple partitioned data set using a partition correction algorithm and redistributing it to the multiple servers to complete the data skew correction. The present invention also proposes a data skew correction device and a computer-readable storage medium. The present invention can realize an accurate and efficient data skew correction function.

This application relates to the field of sensor technology, providing a force sensor and its fabrication method. The force sensor has a disk-shaped structure, including four irregularly shaped elastic beams connected between an outer support ring and a central support platform, and multiple strain gauges on the main surface of the elastic beams. The main surface is parallel or substantially parallel to the disk surface of the disk-shaped structure. Each of the four elastic beams includes four straight arms. The two ends of the first straight arm are orthogonally connected to the second and third straight arms, respectively, and the middle of the first straight arm is orthogonally connected to the fourth straight arm. The fourth straight arm is located on either side of the second and third straight arms. The outer ends of the fourth straight arm and the second and third straight arms are each connected to the central support platform or the outer support ring. Because the irregularly shaped elastic beams include multiple orthogonal straight arms, the strain gauges can be flexibly arranged in a bridging configuration, making the bridging output sensitive only to specific forces or torques, thus improving the decoupling capability of the sensor structure. Simultaneously, placing the strain gauges only on the main surface reduces the mounting difficulty.

The application relates to the technical field of storage computing, and discloses a system, a method and an electronic device for directly searching data in a storage medium, the system comprising a large-scale vector database, a nearest neighbor search operation unit and a host computer, the large-scale vector database is arranged in the storage medium and is used for storing a vector set; the nearest neighbor search operation unit is arranged in the storage medium; the host computer is provided with a CPU and a cache unit; when the nearest neighbor search operation unit receives an NNS_CSD instruction sent by the CPU, the vector of the database is read, similarity measurement is carried out, and the search result meeting the condition is sent to the cache unit, the application directly searches data in the storage medium, the NNS_CSD instruction is used to start the calculation in the storage medium, the nearest neighbor search operation unit in the storage medium is used for vector reading and similarity measurement, compared with a traditional architecture, the migration of data is greatly reduced, and the overall performance is improved.

This invention discloses a maximal clique enumeration method based on MPICH parallel computation for social network streaming data. This method utilizes a structural inversion clique tree (ICTree) to store maximal cliques in the graph. Parallel programming of the algorithm is achieved through the high-performance and widely portable message passing interface MPICH. As the graph structure changes under the streaming data, multiple processes compute in parallel and update the inversion clique tree (ICTree) structure, enumerating maximal cliques in the graph and all maximal cliques in the subgraph formed by nodes within a certain index interval in real time. This invention does not distinguish between edge addition and subtraction changes in the graph during maximal clique enumeration, enabling maximal clique enumeration under mixed edge addition / subtraction updates and large-batch updates. Furthermore, the use of a predefined number of processes for parallel computation solves the real-time processing problem caused by large data volumes, improving the time efficiency of maximal clique enumeration.

The invention provides an animation generation method and device, electronic equipment and a storage medium. The method comprises the following steps: determining an original layer set for generating animations in a to-be-drawn layer set of a current frame; according to the hierarchical relationship of the image layers, each original image layer in the original image layer set is superposed in sequence through downsampling processing to obtain a reference image layer, and the resolution of the reference image layer is lower than that of the image layers in the original image layer set; an animation layer corresponding to the current frame is determined according to the reference layer, and the resolution ratio of the animation layer is higher than that of the reference layer.

The invention discloses a distributed detection method and system for SDN abnormal traffic based on D-VAE-WGAN, and the method comprises the steps: carrying out the cooperative training of a D-VAE-WGAN model based on a collected original traffic sample data set, and obtaining an optimal D-VAE-WGAN model; the cloud server sends a copy of the trained discriminator Dd to each subnet controller, and the copy of the trained discriminator Dd, an original encoder Ge, an original decoder Gd and an original discriminator De in the controller form a detection agent; calculating a reconstruction error and a discrimination error of the reconstructed sample based on the detection agent, and obtaining an error value of the sample; and judging based on the obtained error value of the sample and a set threshold value, and if the error value is greater than the judged threshold value, judging that the sample is an abnormal sample. According to the method, the problem of slow convergence of the GAN is solved by taking the VAE as an encoder, and the risk of'mode collapse 'of the GAN is avoided by taking the Wasserstein distance as similarity measurement. Joint training is carried out by using a generator to exchange parameters, so that each sub-network can learn flow data distribution in other sub-networks, and the overall defense capability of the SDN network is improved.

The invention provides a quantity monitoring method, device and equipment and a readable storage medium. The method comprises the following steps: acquiring channel state information of a space to be measured; performing data processing on the channel state information to obtain first amplitude data of the channel state information; taking the first amplitude data as the input of a target calculation model, and obtaining a monitoring result used for indicating the number of target objects in the to-be-detected space; according to the scheme, the channel state information of the to-be-detected space is obtained, and the features of the target objects can be extracted, so that the number of the target objects passing through the to-be-detected space is determined; by processing the channel state information, the interference of multipath components can be filtered while the characteristics of the channel state information are reserved, so that the calculation pressure of performing calculation according to a target calculation model is reduced. The problems that in the prior art, under the scene that the crossing object density is large and the target objects in the image are seriously overlapped, the image segmentation difficulty is high, and the estimation precision of the number of the target objects is poor are solved.

The present invention discloses a power system peak-shaving demand assessment system, comprising a collection layer, a transmission layer, an edge cloud layer and a management platform; the collection layer is used to collect data; the transmission layer is used to transmit the data collected by the collection layer; the edge cloud layer is used to receive and process data, and the processing of data by the edge cloud layer includes preprocessing and analysis processing; the management platform includes a prediction model, an evaluation model and a visualization module; the present invention quickly processes data through the edge cloud layer to reduce the calculation pressure of the management platform; the management platform includes a prediction model, which predicts power generation through the prediction model, and the management platform includes an evaluation model, which evaluates the peak-shaving process through the evaluation model, and the evaluation and prediction can correspond to each other, so as to facilitate timely regulation when the peak-shaving power is insufficient, and maintain the accurate control of the power in the peak-shaving process.

The invention provides a control method based on a charging station energy management system, and the system comprises an energy management host and a plurality of energy management sub-controllers which are in communication connection with the energy management host. Comprising the steps that S1, the energy management host receives state information sent by the multiple energy management sub-controllers respectively, and the state information comprises the state of charge SOC, the required power Np and the buffer distribution proportion B of the corresponding electric equipment; s2, the energy management host obtains the total power MPW of the charging station, obtains the total power APW of the upper limit value of the energy buffer area distributed to each energy management sub-controller, and calculates the residual distributable power RPW; in the charging peak period, when the multiple sub-controllers carry out rapid power adjustment at the same time, the response time can still be stably kept at the millisecond level, so that rapid suppression can be achieved during instantaneous load fluctuation, and the problems of transformer overload and power distribution imbalance are effectively prevented.

The invention belongs to the technical field of traffic management and image processing, and discloses a traffic accident detection method based on a multi-modal video classification model, which comprises the following steps: S1, detecting and tracking a vehicle; s2, searching a parking vehicle in the current frame; s3, clustering and outputting parking areas; s4, extracting an ROI image sequence of the parking area; and S5, performing video classification by using a multi-mode large model Video CLIP, and judging whether an accident occurs or not. The invention also discloses a traffic accident detection system based on the multi-modal video classification model. The system comprises a video stream acquisition unit, a traffic accident detection unit, an alarm unit and an event retrieval and query unit. The method is high in operation speed, firstly, whether a potential accident exists or not is judged through parking, a potential accident area can be accurately positioned, and the calculation pressure of a follow-up video classification network is relieved; the robustness is better, and the problem of poor model generalization caused by insufficient video training samples can be relieved by utilizing the prior of text information.

The invention relates to the technical field of computer security, in particular to a data storage method which comprises the following steps: S10, identifying and verifying data; the data identification and verification comprises data classification and data verification; s20, encrypting and storing the data; the data encryption storage comprises data encryption and data storage; and S30, a real-time behavior monitoring and tamper-proof mechanism. Through the combination of data encryption and decentralized storage, the risk of data leakage and tampering is effectively reduced; according to the method, symmetric encryption and asymmetric encryption technologies are reasonably combined, and decentralized storage is introduced, so that the calculation pressure in a traditional public key encryption storage scheme is reduced, and the storage efficiency is improved; a distributed storage and redundancy mechanism is adopted, so that the fault-tolerant capability of the system can be improved, and the influence of a single-point fault on data security is avoided; through a multiple backup and redundancy storage mechanism, it is ensured that data can be quickly recovered when attacked or lost.