160 results about "Data efficiency" patented technology

The invention discloses a visual reasoning adjustment method based on mixed training and electronic equipment, and belongs to the technical field of data processing.The method comprises the steps that a visual reasoning data set is determined, and the supervision fine tuning reasoning ability of a visual language model is activated based on the visual reasoning data set, so that the visual language model can complete a target reasoning process; and utilizing a group relative strategy optimization algorithm to optimize the supervised fine tuning reasoning capability of the visual language model. According to the method, the dependence on a large amount of annotated data is reduced, and the data efficiency is improved; and through a dynamic optimization mechanism of reinforcement learning, the method can better adapt to cross-domain tasks, and the generalization ability of the model is remarkably improved.

A system and methods for integrated data processing and protocol adaptation using dyadic distribution-based compression. The system transforms input data into a dyadic distribution, enabling efficient compression through either variational autoencoders or Huffman encoding. A novel protocol appendix generator creates transformation rules for adapting the compressed data to various network protocols. The system interleaves transformation information with the compressed data, enhancing security and ensuring comprehensive data transmission. An enhanced codeword decoder, employing a hybrid neural network architecture, decodes the data and adapts it to target protocols. The system features a protocol handler using meta-learning techniques for adapting to unfamiliar protocols. Continuous learning mechanisms optimize performance over time. This integrated approach offers significant advantages in data efficiency, security, and protocol flexibility, making it particularly suitable for complex, heterogeneous data environments such as IoT networks, cloud computing, and big data analytics.

A system and methods for integrated data processing and protocol adaptation using dyadic distribution-based compression. The system transforms input data into a dyadic distribution, enabling efficient compression through either variational autoencoders or Huffman encoding. A novel protocol appendix generator creates transformation rules for adapting the compressed data to various network protocols. The system interleaves transformation information with the compressed data, enhancing security and ensuring comprehensive data transmission. An enhanced codeword decoder, employing a hybrid neural network architecture, decodes the data and adapts it to target protocols. The system features a protocol handler using meta-learning techniques for adapting to unfamiliar protocols. Continuous learning mechanisms optimize performance over time. This integrated approach offers significant advantages in data efficiency, security, and protocol flexibility, making it particularly suitable for complex, heterogeneous data environments such as IoT networks, cloud computing, and big data analytics.

The invention provides an agricultural robot navigation method based on masked residual Transformer quantum reinforcement learning, and the method is composed of a masked residual Transformer (MRT) network model and the Qua-SAC quantum reinforcement learning supporting the model. The input of the MRT is an original RGB image in a target state under polar coordinates and an original RGB image in a visual state, the feature extraction part is mainly composed of a layer normalization operation layer, a mask multi-head self-attention block, a multi-layer sensor block and a full connection layer, and through the design of the MRT network, target key features can be extracted. The problem of low data efficiency caused by decoupling of a sensing module and target information in an agricultural unstructured environment is solved, and scene representation is promoted to be more focused on target related features. Quantum convolution is introduced in the Qua-SAC quantum reinforcement learning, so that the accuracy and the calculation efficiency of Q value estimation are improved to learn an optimal navigation strategy, and the navigation performance of an agricultural robot in an agricultural non-institution environment is improved.

The invention relates to a meteorological environment information crawling and analyzing method based on a large language model, belongs to the technical field of meteorological data acquisition and analysis, and solves the problems that the efficiency of acquiring meteorological data by depending on manual data search is low, and unstructured data cannot be fully utilized. Comprising the steps of obtaining a meteorological environment information crawler demand, a meteorological environment information analysis demand and a meteorological environment information report generation demand, and constructing a meteorological environment information crawling cue word, a meteorological environment information analysis cue word and a meteorological environment information report generation cue word in a grading manner based on respective preset cue word templates; performing data crawling on the data source website based on the meteorological environment information crawling cue word and a preset meteorological environment information extraction cue word, and performing similarity measurement on a crawling result and an ERA5 data set to obtain meteorological environment information; and generating cue words based on the meteorological environment information analysis cue words and the meteorological environment information report, and performing information analysis on the meteorological environment information to obtain a meteorological environment information analysis report.

The invention discloses an electronic device parasitic parameter network analysis method based on point cloud deep learning and a related device, and the method comprises the steps: obtaining physical characteristic data in a laser-induced quartz material process, and the physical characteristic data comprises laser wavelength, pulse width, photon energy and material characteristics; and inputting the physical characteristic data into the trained neural network model, and predicting the damage threshold of the laser-induced quartz material. According to the machine learning method fusing physical information, unification of data efficiency and physical consistency is achieved, prediction precision and model expandability are remarkably improved, and the method is used for accurately predicting the damage threshold value of the quartz material under the specific laser condition, so that the laser processing technology is optimized, and the processing precision and efficiency of the quartz material are improved.

The invention relates to the technical field of storage, in particular to a storage array controller, a storage array system and a data processing method. The storage array controller comprises a storage interface control module which is used for receiving and analyzing message data sent by a remote host to obtain an operation command, carrying out scheduling management on the operation command based on a namespace and then transmitting the operation command to a redundancy module; the redundancy module is used for performing redundancy calculation on the data, writing the data into a controller memory, and managing the logic address space of the storage array and the controller memory based on the namespace; and the nonvolatile storage control module responds to the data transmission request of the redundancy module and maps the controller memory to realize data interaction with the storage array, so that the problem of low read-write data processing efficiency of the existing storage array controller can be solved.

Imitation learning, or artificial intelligence-based learning from demonstration, aims to acquire an agent policy by observing and mimicking the behavior demonstrated in expert demonstrations. Imitation learning can be used to generate reliable and robust learned policies in a variety of tasks involving sequential decision-making, such as autonomous driving and robotics tasks. However, current imitation learning solutions are limited in their ability to generalize states or goals unseen from the expert's demonstrations. The present disclosure integrates a diffusion model into generative adversarial imitation learning, which, in terms of prior solutions, can provide superior performance in generalizing to states or goals unseen from the expert's demonstrations, provide data efficiency for varying the amounts of available expert data, and capture more robust and smoother rewards.

The invention discloses a new energy data entry method, system and device and a storage medium, and relates to the technical field of data entry, and the method comprises the steps: obtaining machine account data, and obtaining a proportion emphasis parameter; acquiring a period consumption curve, and acquiring a data entry period; recording machine account data of the new energy based on the data recording period; the method is used for solving the problems that in an existing new energy machine account data input method, when machine account data of multiple kinds of new energy is continuously input for a long time, the input period of each machine account data cannot be controlled based on the consumption cost and the consumption proportion of each kind of new energy, and therefore after the machine account data is input, the machine account data cannot be continuously input. The problems that the data processing time is increased and the data efficiency is reduced due to the fact that workers are assisted to carry out effective data analysis are solved.

A problem of imbalanced big data is solved by decoupling a classifier into a neural network for generation of representation vectors and into a classification model for operating on the representation vectors. The neural network and the classification model act as a mapper classifier. The neural network is trained with an unsupervised algorithm and the classification model is trained with a supervised active learning loop. An acquisition function is used in the supervised active learning loop to speed arrival at an accurate classification performance, improving data efficiency. The accuracy of the hybrid classifier is similar to or exceeds the accuracy of comparative classifiers in all aspects. In some embodiments, big data includes an imbalance of more than 10:1 in image classes. The hybrid classifier reduces labor and improves efficiency needed to arrive at an accurate classification performance, and improves recognition of previously-unrecognized images.

The embodiment of the invention discloses a fan control method, device and equipment based on laser radar and a medium. The method comprises the following steps: acquiring radar data corresponding to a plurality of distance sections of a laser radar; fan variable pitch data and fan yaw data are determined according to the radar data; and performing variable pitch control and yaw control on the fan of the laser radar based on the fan variable pitch data and the fan yaw data. According to the technical scheme, the fan variable pitch data and the fan yaw data influencing fan control are determined by using the radar data field output by the laser radar, so that the fan is controlled in an auxiliary manner, and the data efficiency is improved.

The embodiment of the invention provides a transaction data anomaly detection method and device, a program product and a storage medium, and the method comprises the steps: determining a target detection model combination from N groups of anomaly detection model combinations through employing sample transaction data, and enabling the target detection model combination to comprise M target sub-models which are sequentially connected; the first target transaction data is input into a target detection model combination to obtain an anomaly detection result sequence output by the last sub-model in the target detection model combination, in the target detection model combination, the input of the first sub-model is the first target transaction data, and the input of the second sub-model is the second target transaction data; the input of the other sub-models is the first target transaction data and the output result of the previous sub-model; and determining a target anomaly detection result of the first target transaction data based on the anomaly detection result sequence. The problem that abnormal transaction data cannot be efficiently detected in the prior art is solved, and the effect of improving the efficiency of detecting the abnormal transaction data is achieved.

The invention discloses a drug combination patent intelligent identification method and device based on a large language model, and the method comprises the steps: obtaining a standard data set built based on drug combination labeling information of patent data, carrying out the data expansion of the standard data set, and generating an initial training set; a parameter efficient fine tuning technology is adopted, the initial training set is utilized to train the base large language model, and an initial recognition model is generated; screening and identifying difficult samples in the unlabeled patent data by using the initial identification model, and adding the difficult samples into the initial training set to form an enhanced training set; performing iterative training on the initial recognition model by using the enhanced training set and adopting a multi-task learning framework in combination with a comparative learning mechanism to obtain a drug combination patent recognition model; the to-be-recognized patent data is input to the drug combination patent recognition model, and a drug combination information recognition result is obtained.The accuracy, data efficiency and model interpretability of drug combination patent recognition can be remarkably improved, and the computing resource requirement and the manual review cost are reduced.

The invention discloses a VLA task space control method and system for a multi-degree-of-freedom mechanical arm, and relates to the technical field of intelligent robot control, visual information and language instructions when the mechanical arm executes a task are obtained, and a trained VLA model is adopted to obtain a target pose of an end effector in a task space; an IK solver corresponding to the mechanical arm configuration is adopted, and according to the target pose, multiple sets of joint angle solutions of all joints of the mechanical arm are obtained; selecting an optimal joint angle of each joint from the multiple groups of joint angle solutions by adopting a preset optimization criterion; and after the optimal joint angle sequence of the mechanical arm is subjected to track smoothing processing, the mechanical arm is driven to move. Mechanical arm motion control is carried out in combination with a VLA model and an IK solver, the problems that the generalization ability between different mechanical arm configurations is poor, the data collection cost is high and the model interpretability is insufficient are solved, and the generalization ability and data efficiency of mechanical arm control are improved.

The invention provides an autonomous operation obstacle avoidance method and system for a low-altitude aircraft, and the method is executed by the low-altitude aircraft, and comprises the following steps: S1, data acquisition: collecting data through a multi-source heterogeneous sensor group carried by the low-altitude aircraft, airspace control information, real-time meteorological data and collaborative sensing data of other aircrafts sent by the ground sensing base station are received through the air-ground communication interface; preprocessing the collected original data, wherein the preprocessing comprises data format standardization, abnormal value elimination and time synchronization; and S2, data fusion processing: carrying out credibility evaluation on the preprocessed multi-source heterogeneous sensing data and collaborative sensing data. According to the invention, through three-dimensional credibility evaluation and an air-ground data compensation mechanism, the perception precision is greatly optimized, the perception data efficiency is improved in complex environments such as low illumination, strong electromagnetic interference, rainstorm and the like, and obstacle avoidance decision errors caused by data distortion of a single sensor are avoided.

The invention discloses an excavator dynamic load calculation method based on physical information constraint deep learning, belongs to the technical field of load calculation, is used for excavator dynamic load calculation, and comprises the steps that an excavator dynamic model and a neural network model are constructed, and the neural network model comprises an input layer, a conversion layer, a dynamic model learning layer and an output layer; taking the excavator kinetic model as a constraint term of a kinetic model learning layer; the input data of the input layer comprises three angles, angular velocity, angular acceleration and hydraulic pressure variables of the three-section arm, totally 12 variables. According to the method, the sensitivity of the model to unavailable parameters is reduced based on hierarchical configuration of physical constraints, the interpretability of the model is enhanced through mechanical constraint explicit integration exported through a dynamic equation, the data efficiency is optimized by utilizing a dynamic relationship constraint solution space, the network structure design is improved, and the network performance is improved. Dependence on hyper-parameters such as the number of layers and the number of neurons determined manually is reduced, and actual application requirements are better met.

A display driving device includes a first communication circuit which receives first data at a first data rate via a communication line and determines whether the first data has an error according to a first rule; a second communication circuit which receives second data via the communication line at a second data rate which is higher than the first data rate, and which determines whether the second data has an error according to a second rule which is different from the first rule; and a circuit that controls pixels on a display panel to be driven in accordance with image data included in the second data. driving a display panel, thereby improving the accuracy and efficiency of data validity verification by verifying data efficiency in data communication in different ways depending on the type and operation of the sent / received data.

The invention relates to the technical field of storage, in particular to a cache management method, a storage array controller, a device, equipment and a medium. The method comprises the following steps: distributing a cache region for an access command in a static cache subspace according to a namespace to be accessed by the access command; receiving, reading and writing target data of the storage array based on the access command, and caching the target data to the cache region; wherein the static cache subspace is obtained by dividing a static cache space in the controller memory according to the namespace. Therefore, the problem that the efficiency of reading and writing the storage array data is low due to unreasonable cache space distribution can be effectively solved.

The invention discloses a multi-source heterogeneous knowledge injection type prompt learning method for legal criminal name prediction. The method comprises the following steps: step 1, obtaining and coding fact elements; 2, knowledge matching and prompt injection; 3, reasoning the legal language model; and 4, legal crime name category mapping. The method has the advantages that prediction accuracy is improved; the problem of input length limitation is solved; the field adaptability is enhanced; the data efficiency is improved; the interpretability of the model is enhanced; multi-task learning is supported; and intelligent development of laws is promoted.

Systems, devices, and methods including an aerial vehicle having a global positioning system (GPS) and at least one trace-gas sensor configured to generate gas data; and a processor having addressable memory, the processor configured to: determine a flight envelope based on a received spatial location, a received spatial location of the one or more potential gas sources, a received desired level of confidence, and a received wind data; determine a flight path for the aerial vehicle, where the flight path covers a portion of the determined flight envelope; and determine based on a received gas data whether a gas leak is present in the received spatial location to the received desired level of confidence.

The invention relates to the technical field of communication, in particular to a data reading and writing method, system and equipment based on an I2C bus. The method comprises the following steps: filling a target register address in a register field in a preset I2C data frame structure; according to the number of the slave devices to be subjected to batch read-write operation, setting device data units in the I2C data frame structure, wherein the number of the device data units is equal to that of the slave devices; each equipment data unit is filled with an equipment ID of the corresponding slave equipment and a corresponding operation identification value; the operation identification value is used for indicating the corresponding slave equipment to determine the opportunity of executing the read-write operation; according to the filled I2C data frame, an I2C data message corresponding to the read-write operation is obtained, and the I2C data message is sent to the I2C bus, so that each slave device reads and writes data of the slave device according to the I2C data message, and the problem that the I2C data read-write efficiency is low can be effectively solved.

The invention provides a multi-controller data writing method and device, electronic equipment and a storage medium, and the method comprises the steps: receiving a data writing request, and determining to-be-written data corresponding to the data writing request and a target data strip in which the to-be-written data is to be written into an RAID (Redundant Array of Independent Disks); accessing metadata in the associated memory, and determining a target value of a bit corresponding to a target data strip in the metadata; and based on the target numerical value, determining whether the target data stripe is occupied, and determining whether to execute a data writing operation corresponding to the to-be-written data. The memory resource sharing among the memory controllers is realized through the memory consistency interconnection among the plurality of memory controllers, and the plurality of memory controllers are directly associated together through the interconnection memory, so that the serial connection of the physical level is realized. According to the technical scheme, the mode that the multiple storage controllers jointly operate the same part of metadata in the interconnected memory is adopted, a stripe lock mechanism is cancelled, the IO writing processing time delay is reduced, and the data writing efficiency is improved.

The invention discloses a system and method for improving data configuration efficiency of a test machine, and relates to the technical field of board card testing, the system comprises an FPGA PCIE module, a DMA address data analysis module and a data transmission module, a general data frame is divided into an independent 32b address field and an independent 32b data field through the arranged DMA address data analysis module, then bit [31] and 0 of an 32b address are analyzed to represent write operation, and the data transmission module is connected with the FPGA PCIE module. 1 represents read operation, an identifier is independently extracted and embedded into a subsequent instruction packet, then a bit [27: 0] of a 32b address is extracted, a target daughter card number is determined according to a preset address mapping table, and data is packaged after determination, so that data distribution of different address fields is realized, the data transmission module receives the data and then sends the data to each post-stage daughter card module in parallel, and data read-write is realized. Therefore, the read-write speed of the test machine is greatly increased, and the read-write efficiency is further improved.

The invention discloses a report generation method based on a large model, electronic equipment and a computer program product, and relates to the technical field of artificial intelligence, the report generation method comprises the following steps: identifying a logic structure, a placeholder and a task dependency relationship in a preset report template, and initializing a context state carrier based on the logic structure, the placeholder and the task dependency relationship; obtaining a report generation requirement, converting the report generation requirement into SQL query data, and generating a cue word based on the SQL query data; and inputting the cue word into a preset generation model to obtain filling content, updating the filling content to the context state carrier, and filling placeholders of the report template based on the context state carrier to obtain a final report. According to the method, the defect of low efficiency of manual data arrangement is overcome, and the problem that a simple template filling tool cannot process a complex logical relationship is solved, so that the report generation efficiency is improved.

The present disclosure relates to a method and apparatus for realizing cross-instance routing of a structured query statement, and a system. The method comprises: a first database service instance end receiving a structured query statement; the first database service instance end determining whether the structured query statement has the capability to be sent to a second database service instance end; and when it is determined that the structured query statement has the capability to be sent to the second database service instance end, the first database service instance end establishing a connection to the second database service instance end, and forwarding the structured query statement to the second database service instance end for execution. The present disclosure solves the technical problem in the relevant art of the low efficiency in cross-instance data access.

The invention discloses a construction site worker behavior intelligent identification and early warning system, which collects multi-directional video streams in real time through a front-end video collection module, guarantees timely data updating, and facilitates real-time monitoring of worker behaviors. The early warning and display module displays a field picture, a worker skeleton diagram, a behavior state and duration time in real time, early warning is immediately carried out when abnormity occurs, an instant feedback mechanism of the early warning and display module enables safety management personnel to rapidly know the abnormity and take measures, accidents caused by information lag are avoided, and the defect that the real-time early warning capability is insufficient is overcome; the behavior recognition module adopts an EnhanceST-GCN model, can better capture a space-time relation of skeleton key points, accurately recognize a complex behavior mode, improve recognition precision and solve the problem of low recognition precision, and meanwhile, the model is high in video data processing efficiency, rapid in analysis, feature extraction and classification, reduces unnecessary calculation steps and meets the real-time monitoring requirement of a construction site.

The invention relates to the technical field of computer vision and medical instrument management, and discloses a medical instrument intelligent identification method, device and equipment and a medium for solving the problem that the accuracy of surgical instrument identification is insufficient. And a dual-polarization acquisition and pixel difference technology is combined to accurately separate a mirror surface highlight area from a diffuse reflection area, so that the image definition is improved. Through the first analysis model, the determined and uncertain identification areas are rapidly determined, and the second image of the uncertain area is focused and complementarily collected, so that redundant collection is reduced, and the data efficiency and accuracy are improved. The image subjected to glare-removing tuning processing has clearer feature information, and the recognition accuracy of the uncertain area can be remarkably improved after re-analysis. And finally, fusing the first and second determination identification results to comprehensively improve the accuracy and reliability of medical instrument identification, thereby effectively solving the technical problem of insufficient accuracy of existing medical instrument intelligent identification.

The utility model relates to the technical field of data acquisition, and discloses a vehicle-mounted data acquisition device which comprises a shell, the top of an inner cavity of the shell is movably sleeved with a first steering shaft, a first motor is fixedly installed on the top of the inner cavity of the shell and located on the left side of the first steering shaft, and the bottom end of an output shaft of the first motor is fixedly sleeved with a first gear. The right side of the first gear extends outwards and is meshed with a second gear, the second gear fixedly sleeves a first steering shaft, a rotating groove is fixedly formed in the bottom end of the first steering shaft, a second motor is fixedly installed at the top of an inner cavity of the rotating groove, an output shaft of the second motor is fixedly sleeved with a third gear, and the lower portion of the third gear extends outwards and is meshed with a fourth gear. The two sets of motors are matched, so that the camera shooting angle of the device is flexibly adjusted, the number of view blind areas is small, the flexibility of the device is improved, the data collection efficiency of the device is improved, the camera shooting device is rapidly mounted and dismounted, maintenance is more convenient, and the labor intensity of workers is relieved.

The present application relates to a magnesium alloy ingot stacking robot, which belongs to the field of ingot stacking robots, and includes a mechanical arm, the end of which is connected to a support plate, and two opposite side walls of the support plate are connected to horizontal clamping plates for clamping the magnesium ingots, and each of the horizontal clamping plates is connected to a cylinder for driving the horizontal clamping plates to clamp the magnesium ingots on one side close to the support plate, and a linkage component is connected to one side of the cylinder, and the linkage component is used to drive the piston rod of the cylinder to move. The present application has the effect of improving the stacking efficiency of magnesium ingots.

The invention provides a method and device for calibrating test data in batches and a computer readable storage medium. The method comprises the steps that a test case set is obtained, calibration variables and calibration values of the test case set are recognized, the calibration variables and the calibration values are stored in a variable array, and the test case set is an executable automatic test case data set; determining a calibration variable address according to the variable name of each calibration variable, and determining a partition of each calibration variable according to each variable address; and performing calibration processing on each calibration variable according to the partition and the calibration value of each calibration variable so as to complete calibration of the test data. The problems of low efficiency and high error rate of manual data calibration in the prior art are solved.