18results about How to "Reduce decision time" patented technology

The invention discloses a VVC coding unit rapid division method and system based on image frequency domain characteristics, and the method comprises the steps: employing a first image frequency domain characteristic, a first classification label and a first weight to train a group of binary classifiers based on a lightweight gradient lifting model, and obtaining a group of initial binary classifiers; performing feature selection through the group of initial binary classifiers, and training the group of initial binary classifiers by adopting the selected features to obtain a group of dimensionality reduction binary classifiers; performing parameter adjustment on the group of dimensionality reduction dichotomous classifiers to reduce the scale of the group of dimensionality reduction dichotomous classifiers to obtain a group of division decision classifiers; initializing online sample data, and deciding the division mode in each division decision classifier according to the online sample data to obtain a group of target division modes; and quickly partitioning the VVC coding unit according to a group of target partitioning modes. According to the invention, the division efficiency and decision accuracy of the coding units are improved, the video quality and code rate are ensured, and the coding time is shortened.

The invention discloses a viewpoint planning method based on deep reinforcement learning. The method comprises the following steps: acquiring a sheltered image and a camera attitude and obtaining state information; inputting the state information into a trained DQN network to obtain a camera motion action at the next moment; and controlling the mechanical arm to move to adjust the camera position. In order to improve the training efficiency of the DQN network, a training strategy of global trend guide learning is creatively provided. The global trend guided learning strategy refers to a method for forming a global trend by recording viewpoint areas of successfully detected fruit stems in the same scene in a learning process, and then guiding a network learning process by using the global trend. According to the invention, viewpoint planning is realized by a brand-new technical means, so that the problem of shielding of the grape picking robot in the picking operation process is solved. According to the method, the viewpoint planning time is greatly shortened under the condition that the high picking success rate is kept, and a new technical scheme is provided for the field of active vision technology application.

The invention relates to the technical field of resource transmission configuration, and discloses a transmission configuration method and system for medical air transportation, and the method comprises the steps: obtaining user data and resource data through a scene modeling module, constructing a task scene model, and providing a basis for subsequent decision making; the post-sending decision-making module generates a post-sending decision-making strategy based on a result of the task scene model to obtain a post-sending scheme, and selects the post-sending scheme according to an actual situation, so that the decision-making accuracy and the resource utilization rate are improved; the simulation deduction module performs simulation deduction according to the task scene model and the selected post-delivery scheme, simulates the medical air delivery post-delivery process, and visually detects deduction data in the simulation deduction process; and after simulation deduction is finished, the decision evaluation module evaluates deduction data obtained by visual detection to optimize a post-delivery decision strategy, so that the delivery configuration system can quickly and effectively respond to a medical air delivery post-delivery scene of batch sick and wounded in wartime, and the emergency response capability of the delivery configuration system is improved.

The present application relates to the technical field of smart city, especially relates to a smart city bus stop board dynamic information accurate pushing method and system, wherein a query semantic representation structure is constructed by collecting query requests and scene state information, a joint understanding representation is obtained by modeling the scene state information multi-level semantic association, scene constraint multi-hop reasoning is executed in a bus knowledge association network, a candidate answer sequence is generated by filtering, an optimal answer is selected by calculating a scene adaptation score, pushing content is generated, interactive feedback is collected, feature fusion weights are adaptively adjusted to form a closed loop optimization, and accurate dynamic pushing of bus information and user experience improvement are realized.

The invention provides a flexible production management method and system for multi-variety small-batch production orders. The method comprises the following steps: when a new order is received by using an SAP system, obtaining order adjustment information; inputting the order adjustment information into a decision recommendation model to obtain decision recommendation information; wherein the decision recommendation model is integrated in an MES system, and the decision recommendation information comprises configuration parameter information of a production line; adjusting a production plan according to the decision recommendation information, and performing flexible production according to the adjusted production plan; wherein the order adjustment information comprises new order information, production line information and material management information of the WMS. Through analysis and deep learning of production big data, the decision recommendation information is automatically generated for multi-variety and small-batch production characteristics and frequent change of customer orders, so that the production plan is adjusted, the production efficiency is improved, and the purpose of intelligent flexible production is achieved.

The invention provides a plant protection unmanned aerial vehicle path planning method based on reinforcement learning. The plant protection unmanned aerial vehicle path planning method comprises the following steps: acquiring a starting point, a target point and static obstacle position information; establishing a two-dimensional map to obtain a line patrol path distributed in a reciprocating bow shape, and decomposing the line patrol path into a plurality of sub-target tasks; establishing an unmanned aerial vehicle model, setting an unmanned aerial vehicle initial state s and a proper reward function r based on a near-end strategy optimization algorithm, and obtaining rewards through interaction with the environment; updating the state of the unmanned aerial vehicle and taking the next action; when the unmanned aerial vehicle encounters an obstacle, the unmanned aerial vehicle completes an obstacle avoidance task according to an obstacle avoidance principle and then returns to a line patrol task, and completes a sub-target tendency task according to an algorithm process; after all obstacle avoidance behaviors are completed and n sub-target tasks are approached, the unmanned aerial vehicle arrives at a terminal point E; the optimal path planning result is provided for the unmanned aerial vehicle by accumulating the reward value, the spraying operation efficiency of the unmanned aerial vehicle is improved, the operation safety of the unmanned aerial vehicle is considered, and the labor cost is effectively reduced.

The application relates to the fields of green low-carbon technology and construction technology, and discloses a shield tunnel-based repair scheme decision method, device, equipment and medium, which comprises the following steps: generating a stiffness recovery index of a shield tunnel under each repair scheme through a preset stiffness recovery model; generating an average toughness index of the shield tunnel under each repair scheme according to the stiffness recovery index of the shield tunnel under each repair scheme, a stiffness degradation coefficient of the shield tunnel under each repair scheme, a design life of the shield tunnel under each repair scheme and a preset toughness model; generating a comprehensive evaluation value of each repair scheme according to the average toughness index of the shield tunnel under each repair scheme, a carbon emission amount corresponding to each repair scheme and a preset comprehensive evaluation model; and selecting a repair scheme with the maximum comprehensive evaluation value as an optimal repair scheme of the shield tunnel. The application is beneficial to improving the decision efficiency of the optimal repair scheme of the shield tunnel.

The invention relates to the technical field of tourism management, and particularly discloses a digital cultural tourism management system based on multi-source data analysis, which comprises a data acquisition module, a resource integration module, a travel route generation module and a dynamic information updating module, the data acquisition module is used for collecting cultural tourism data from various data sources and storing the cultural tourism data into a database; the resource integration module is used for integrating various culture and tourism resource information and updating the resource information regularly; the travel route generation module is used for intelligently planning a travel route according to tourist demands and scenic spot characteristics and generating a recommended route; for diversified cultural tourism data, an entity network fusing heterogeneous information is established, user travel itinerary information, various tourism resources, scenic spot characteristics and related cultural information are integrated together, and by means of an intelligent route planning function, the system can plan the travel itinerary information according to personal preferences, departure places and time arrangements of tourists. And generating personalized tourist route recommendation.

The application uploads the data obtained by the vehicle-mounted camera sensor of the unmanned mine car to the cloud server, carries out feature recognition on the cloud data through the cloud server, then selects the road curvature, the slope, whether there is an obstacle in the front road, whether there is a fork in the front road as the characteristic parameters according to the characteristics of the mine area to realize the data screening of the vehicle-mounted camera, so that the data input into the decision model of the unmanned mine car is reduced, and at the same time, the speed information of the unmanned mine car is referred to in the vehicle-mounted camera data screening process, and the screening weight is dynamically determined, so that the decision time of the decision model is reduced, and according to the characteristics of the mine area, the sensor data screening and the dynamic determination of the screening weight are considered, most of the interference garbage data is considered, the decision time is reduced, and the decision accuracy of the decision layer is improved.

The application discloses a multi-stage peak position detection method for satellite communication and relates to the field of satellite communication, which comprises the following steps: dividing a joint / complete three-dimensional search space into a plurality of low-dimensional independent search spaces; performing peak position detection in each independent search space by using multi-stage step-by-step decreasing steps; and performing angle jumping during the peak position detection process to avoid the influence of a power dead zone, so that beam alignment of a target satellite is completed while the total search number is reduced.

The present invention provides an information-based medical prevention decision system, which belongs to the field of intelligent decision technology and includes: a vector comparison module, which is used to obtain the detection information of target patients under different medical subjects based on an information platform, and standardize the detection information under each subject respectively and combine the set judgment vector under the corresponding subject to obtain the examination result of the corresponding subject; a measure acquisition module, which is used to obtain preventive measures consistent with each examination result; a subtree construction module, which is used to obtain several combination results according to the subject combination method and combine the preventive measures to construct several sub-decision trees; a combined tree construction module, which is used to set link lines to all sub-decision trees according to the prevention association relationship and treatment association relationship between each combination result to obtain a combined decision tree; a decision analysis module, which is used to perform final decision analysis on the combined decision tree and output a medical prevention decision. The decision time is effectively saved and the decision efficiency is improved.

The present application relates to the technical field of logistics distribution, and discloses a package delivery method and device, computer equipment and storage medium. The package delivery method determines a delivery starting point, a target delivery community, and package size parameters and package waybill addresses of each package to be delivered according to a batch package delivery request; performs clustering analysis on all the packages to be delivered according to the package waybill addresses, and obtains a plurality of package sets; filters candidate delivery spaces of each package set according to set volume parameters; for each package set, determines a delivery comprehensive score of each candidate delivery space according to the delivery starting point, set volume parameters, and spatial attribute information of the candidate delivery space, and determines the candidate delivery space with the highest delivery comprehensive score as the target delivery space of the package set; and generates package set delivery tasks corresponding to each package set based on the target delivery space and sends them to the delivery party, thereby improving the rationality and delivery efficiency of package combination delivery.

The utility model relates to the field of household daily necessities, and provides an auxiliary positioning early warning device for a drilling platform. The auxiliary positioning early warning device for the drilling platform comprises a rectangular long frame, a flooding detection mechanism, a GPS positioning module, a 4G transmission module, a controller and an alarm mechanism, and the front end face and the rear end face of the lower end of the rectangular long frame are fixedly sleeved with U-shaped mounting plates. According to the auxiliary positioning early warning device for the drilling platform, through cooperation of the floating ball and the waterproof pressure sensor, instantaneous wave interference is avoided, the water logging condition can be accurately detected, the false alarm problem of a traditional water level sensor is avoided, meanwhile, the position of the platform is monitored in real time in combination with the GPS positioning module, and the positioning accuracy is improved. Once a safety range is exceeded, offset early warning is triggered immediately, simultaneous operation of multi-dimensional safety monitoring, flooding alarm and offset early warning is realized, risks can be analyzed in a linkage manner, whether water enters due to drifting or not is analyzed, decision time is shortened, and emergency response can be optimized.

The invention relates to the technical field of intelligent decision making, and provides a combat decision making method and device for a combat entity, a computer readable storage medium and electronic equipment, the surrounding space of the combat entity is divided into a plurality of defense areas, and N combat weapons are deployed on the combat entity, each combat weapon is deployed in at least one defense area on a combat entity, and the method comprises the following steps: obtaining feature parameters of M incoming attacking targets for the combat entity at the current moment, and obtaining overall combat effectiveness parameters of the combat entity for the M incoming attacking targets at the current moment; inputting the overall combat effectiveness parameters and the characteristic parameters into a trained combat decision model, and deciding a combat strategy for M incoming attack targets through the trained combat decision model; wherein the combat strategy is used for representing the attack probability of each combat weapon for each incoming attack target. The method can effectively cope with a multi-target and dynamically changing complex battlefield environment.

The invention relates to the technical field of data processing, and discloses an accurate flow pushing method and system based on user behavior prediction and a storage medium, and the method comprises the steps: obtaining multi-dimensional behavior data of a user, carrying out the preprocessing of the data, obtaining a multi-dimensional feature vector, inputting the multi-dimensional feature vector into a deep learning model and a Bayesian network, and generating user behavior prediction data; performing type sorting on advertisement traffic available to be sent based on the data; pushing parameters are calculated based on DSP real-time bidding data and budget data, the similarity between advertisement types in pushed advertisement flow and the pushing parameters is calculated, and priority distribution and putting are carried out on the advertisement flow according to the similarity; after delivery, delivery feedback data is generated; putting feedback data of the DSP are collected, and advertisement feedback information after advertisement flow is put is obtained from the putting feedback data; and adjusting parameters in a similarity algorithm based on the advertisement feedback information. According to the invention, accurate delivery of DSP advertisements is realized through a more intelligent traffic prediction and push method.

The application relates to the technical field of fire safety, in particular to an automatic water-spraying fire extinguishing system and method based on multi-modal perception and intelligent decision-making, which comprises a perception module, a data processing and control module and an execution module; the data processing and control module is embedded with a hybrid deep learning model; the hybrid deep learning model fuses a convolutional neural network for analyzing smoke images and a long short-term memory network for analyzing temperature sequences; multi-modal sensor data can be deeply fused and intelligently understood, so that high-accuracy fire identification can be realized within 3 seconds.The beneficial effects are that, through deep fusion and interaction of hardware sensors and intelligent algorithms, the problem of slow response and high false alarm rate of traditional systems is effectively solved, and the system is particularly suitable for prevention and control scenes of new fire risks such as electric bicycles, new energy vehicles and the like.

The invention discloses a store inventory dynamic management and sales demand adaptation optimization method, and belongs to the field of data processing and supply chain optimization. According to the method, historical sales data and cost parameters of all stores are centrally collected and cleaned; based on the cleaned data, calculating a sales rhythm factor K used for representing sales volatility and a demand balance factor M used for representing demand difference between the stores; dynamically correcting the calculation result of the classic economic ordering batch EOQ model by using K and M to obtain the optimal purchase batch of the central warehouse and the optimal replenishment batch of each store; furthermore, the overall optimization of the inventory cost and the response efficiency is realized based on a standardized split inventory unit, a priority distribution unit, real-time monitoring and intelligent allocation decision and parameter closed-loop optimization. According to the method, the problems of high inventory cost and poor adaptability caused by model stiffness, poor data quality and response lag of a traditional inventory management method are effectively improved.