13results about How to "Strong coping ability" patented technology

The invention discloses an online public opinion research, judgment, analysis, classification and grading early warning method and system, and the method comprises the steps: constructing a malicious public opinion data detection and judgment model, and carrying out the iterative training of a pre-training neural network through large-scale online public opinion historical data, so as to achieve the fine adjustment and optimization of the model; after malicious public opinion data are detected, fine-grained classification of malicious public opinions is realized through deep semantic comprehension and multi-dimensional text structure analysis by adopting BERT and ERNIE deep learning models and combining semantic role labeling and dependency syntactic analysis technologies; a malicious public opinion data grading and distinguishing model is constructed, deep correlation between public opinion events is captured through multi-layer convolution operation, and the recognition capacity for public opinion sensitivity and urgency degree is improved. According to the invention, malicious public opinion data related to a specific industry field can be efficiently and accurately identified and classified, and the malicious public opinion data is subjected to severity grading, so that the accuracy and the real-time performance of network public opinion management are improved.

The invention relates to the field of audio signal processing, in particular to a sperm whale cry enhancement method based on a VQ-MAE network, which comprises the following steps: step 1, performing noise reduction processing on a sperm whale cry audio signal, and adjusting the amplitude range of the audio signal; 2, processing a frequency spectrum image, and taking an extracted feature vector as a part of MAE network input; 3, processing the audio signal through an LSTM network, and extracting the time sequence characteristics of the original audio signal; 4, designing and training a multi-mode mask encoder (MAE) network, and training fusion weights of different modes; and step 5, performing masking, and selecting a strategy training network with the best masking effect to perform data enhancement. According to the method, the time domain, frequency domain and spatial domain features of the multi-modal network fusion sound signals are introduced innovatively, semantic details in the wry of the sperm whale are captured and enhanced more accurately and deeply, and powerful support is provided for classification and recognition of the wry.

The invention provides a vision-based collaborative robot safety control method, which comprises the following control steps: S1, acquiring scene information of a collaborative robot, and acquiring a target coordinate and a pixel position based on the scene information; and S2, the specific position of an obstacle around the target and a surrounding three-dimensional scene are recorded based on a camera on the collaborative robot, and the three-dimensional scene is modeled and projected on an operation screen, so that the collaborative robot can be ensured to achieve the effects of rapid positioning, moving and grabbing through mutual cooperation of a plurality of modules; meanwhile, a potential field method is adopted for route planning, so that the collaborative robot can be prevented from colliding with an obstacle during planning movement, it is ensured that the whole collaborative robot has good environment sensing and coping capacity, and the overall efficiency in the man-machine collaborative operation process can be effectively guaranteed; and the robot can be applied to any man-machine cooperation production environment without excessive adjustment, so that the development of the field of cooperation robots is promoted.

The invention provides a driving behavior decision planning method, system and device, and a storage medium, and the method comprises the steps: constructing a state space according to the driving state of a current vehicle, and constructing an observation space according to a plurality of state spaces and environmental factors; constructing an action space according to the current stress conditions of front and rear wheels of the vehicle, and constructing a reward function according to behavior prediction, emergency handling and riding comfort in the historical driving data; the state space and the observation space are changed by executing the action space, and a reward value is given to the change of the state space and the observation space through the reward function so as to train each action space; and in response to completion of training, determining a maximum reward value according to the current state space and the current observation space, and executing an action space corresponding to the maximum reward value to realize decision planning of the driving behavior. The method can make a safe and reasonable decision about the shortest distance to the target location, is high in capability of coping with emergencies, and improves the riding comfort of passengers.

The invention discloses a supply chain management optimization method and system based on a large model. The method comprises the steps of data acquisition and processing, large model demand prediction, inventory management optimization, supply chain operation optimization and supply chain management optimization. The invention relates to the technical field of supply chain management, in particular to a supply chain management optimization method and system based on a large model, according to the scheme, process design of demand prediction, inventory management optimization and operation optimization is sequentially carried out, and complex integrated supply chain data can be better processed; a long-short-term memory neural network combined with a transformer model is adopted to carry out large model demand prediction, and the processing performance of time sequence trend data is optimized; inventory management is carried out by adopting a deep enhanced hybrid linear programming method combined with graph convolution enhancement, and the extraction quality of transport dynamic edge features is improved by adding an aggregation embedding mode of node features and edge features; and a generation bidirectional encoder combined with a transformer model is adopted, so that the intelligence of the method is improved.

The invention discloses a big data method for emergency response and disposal of water pollution accidents. The method comprises the following steps: deploying various types of sensor networks in a pollution-prone area; transmitting the real-time data acquired by the sensor to a cloud storage system; monitoring a real-time event mode in the environmental data by using a complex event processing technology, and generating a real-time environmental data model; a machine learning technology and a geographic information system are used to analyze real-time environment data to predict a diffusion path and an influence range of pollutants, and a pollution source position is rapidly positioned; and intelligent emergency response measures are implemented through the intelligent decision support system. A real-time data stream processing technology and complex event processing are used, environmental data can be monitored in real time, emergencies can be quickly responded, the response time can be greatly shortened due to the real-time performance, pollution events can be found and processed earlier, and therefore environmental damage and influences of a water purification system are reduced.

The invention discloses an air combat deep reinforcement learning hierarchical algorithm based on a time scale separation theory, which combines long-period decision and short-period control, and realizes unmanned aerial vehicle autonomous air combat maneuver decision optimization through a two-stage training strategy. Firstly, the state quantity of the unmanned aerial vehicle is separated into a long-period state quantity and a short-period state quantity according to the time scale, the long-period state quantity comprises the position and the relative speed, and the short-period state quantity comprises the roll angle speed, the pitch angle speed, the yaw angle speed and the like. The long-period decision is optimized through a PPO algorithm, and expected values of short-period control, such as an expected roll angle, a pitch angle and a speed, are output; and the short-period control realizes accurate adjustment of the attitude of the unmanned aerial vehicle through a PID controller. The algorithm adopts two-stage training, in the first stage, convergence is accelerated through a single-step reward strategy, and in the second stage, combat performance is optimized through a final reward. The maneuverability and autonomous decision-making ability of the unmanned aerial vehicle in a complex air combat environment are improved, and the method has the characteristics of quick response and efficient learning and is widely applied to autonomous air combat tasks of the unmanned aerial vehicle.

This invention discloses a microcapsule-modified self-healing mortar, its preparation method, and its application, belonging to the technical field of stone cultural relic protection and restoration materials. The microcapsule-modified self-healing mortar, by mass percentage, comprises the following dry materials: 50%-70% hydraulic lime, 12%-24% nano-kaolin, 15%-25% quartz sand, 0.05%-0.2% additives, and 0.25%-2% microcapsule powder; wherein the core material of the microcapsules in the microcapsule powder is tung oil, and the wall material is barium alginate. The microcapsule-modified self-healing mortar possesses high weather resistance, excellent compatibility, and long-term durable self-healing function. By introducing tung oil-barium alginate microcapsules, the microcapsule-modified self-healing mortar synergizes with traditional restoration mortar raw materials, enabling the mortar material to trigger active repair components in the early stages of deterioration, achieving self-healing and self-repair effects.

The present invention provides a visible light-thermal infrared target tracking method based on robust spatiotemporal context modeling of a state space model, which belongs to the fields of machine learning, computer vision, and target tracking. By utilizing the linear long sequence modeling capability of the visual state space model, long-term cross-frame integration components and short-term historical trajectory cues are embedded into a bidirectional visual state space model encoder (Mamba), and unified spatiotemporal context modeling of global appearance information and local motion information is performed in a single-stream framework. The present invention breaks away from the limitation that the computational complexity of the attention mechanism grows with the square of the sequence length, and for the first time proposes a visible light-thermal infrared tracking framework based on state space models (State Space Models). The method of the present invention utilizes the advantage that its computational complexity grows linearly with the sequence length, and can achieve efficient operation under multi-source long sequence input with lower computational complexity.

The invention discloses an integral series and gain setting method and device based on snowfall data. According to the method, historical snowfall data, multi-band reflectivity data and forecast data of a meteorological station are obtained, a regression model is established to calculate the relation between snowfall and reflectivity changes of different bands, then the relation is utilized to correct an integral series and gain growth model, and a lookup table is established. And on the basis of the forecast data, dynamically calculating the correction value of the integral series and the gain, and finally, comprehensively calculating according to the basic reflectivity data and the correction value to obtain the adjusted integral series and gain setting. The method improves the precision and reliability of remote sensing observation data, and is especially suitable for remote sensing observation tasks under complex meteorological conditions such as snowfall.

The invention provides a heat management system which comprises a refrigerant circulation subsystem and a secondary refrigerant circulation subsystem, the secondary refrigerant circulation subsystem comprises a hot fluid circulation loop and a cold fluid circulation loop, the heat management system further comprises a plurality of heat exchange load tail ends, and when the heat exchange load tail ends are in a heating state, the heat exchange load tail ends are in a heat exchange state. An inlet and an outlet of each heat exchange load tail end are communicated with the hot fluid circulation loop so that the heat exchange load tail ends in the heating state can be connected in parallel. An inlet and an outlet of each heat exchange load tail end are communicated with the cold fluid circulation loop so that the heat exchange load tail ends in the refrigeration state can be connected in parallel. The state of each heat exchange load tail end can be independently adjusted, so that the functional requirements of each temperature adjusting object are met, and the adverse effect of fluctuation of one heat exchange load tail end on other heat exchange load tail ends is reduced.

The invention discloses a cross-platform virus and Trojan horse detection method and system based on deep learning, and the method comprises the steps: scanning codes, files and processes in a system, obtaining each node and a dependency relationship thereof, and constructing a propagation path dependency graph; modeling the constructed dependency graph by using a graph neural network, and dynamically predicting a potential remote propagation path in combination with an iterative update and convergence mechanism of a node feature vector; through real-time monitoring of communication behaviors and network traffic between processes, the system can adaptively adjust the propagation path dependency graph, preferential isolation and early warning are carried out on paths with high propagation risks, and the protection capability of the system is significantly improved.

The utility model discloses a continuous desolventizing device for glyphosate production, which comprises a first-stage desolventizing device, a second-stage desolventizing device and a third-stage desolventizing device which are sequentially connected in series and respectively perform step-by-step desolventizing at different temperatures, the first-stage desolventizing device, the second-stage desolventizing device and the third-stage desolventizing device adopt desolventizing kettles with the same structure to carry out desolventizing work; the first-stage desolventizing device, the second-stage desolventizing device and the third-stage desolventizing device are respectively connected with a first-stage condenser, a second-stage condenser and a third-stage condenser; the first-stage condenser, the second-stage condenser and the third-stage condenser are connected with a tail gas absorbing device and a collecting tank; a methylal online analyzer is arranged on the primary desolventizing device, and a methanol online analyzer is arranged on the secondary desolventizing device; generally, the glyphosate desolventizing device has the advantages of high desolventizing efficiency, high automatic detection and analysis degree, reduction of side reactions and improvement of glyphosate product quality.