401 results about "Operation scheduling" patented technology

The invention discloses a power transmission line fault diagnosis and operation and maintenance scheduling method and system based on networking learning, and relates to the technical field of power transmission line fault diagnosis operation and maintenance scheduling. Related data is extracted to construct a high-risk equipment area and a visual high-risk area thermodynamic diagram, a visual risk grading diagram is constructed in combination with electrical quantity data, and meanwhile, an intelligent recognition storage network and a fault type classification recognition model are constructed in combination with a convolutional neural network-long and short-term memory network hybrid model; the model is optimized through networking learning and an attention mechanism, maintenance teams and resources are autonomously allocated in combination with an operation and maintenance management system, then autonomous optimization and closed-loop operation are achieved, full-process coverage of fault sensing, intelligent decision making and efficient response is achieved, the response time after a line fault occurs is remarkably shortened, and the maintenance efficiency is improved. And the fault handling and operation maintenance capabilities of the power grid system are comprehensively enhanced.

The invention discloses an intelligent electric equipment monitoring and optimizing method, and relates to the technical field of electric power, and the method comprises the steps: collecting high-dimensional voltage-current time sequence data and transient event marks, calculating topological invariant features, inputting the topological invariant features to a lightweight neural network model, and recognizing the features of all electric equipment; performing abnormal attribution and anti-fact energy efficiency prediction by using causal reasoning and dynamic regularization regression based on the identified characteristics of each electric device, and constructing a multi-objective optimization function through the abnormal attribution and anti-fact energy efficiency prediction; and based on the multi-objective optimization function, generating an equipment operation scheduling strategy through a deep reinforcement learning agent, based on the operation scheduling strategy, sending a control instruction to the electric equipment, and collecting an operation result feedback in real time for optimization and updating. According to the method, through fusion of topological features, causal reasoning and safety reinforcement learning, the precision, robustness and safety of monitoring and optimization of the intelligent electric equipment are improved.

The invention provides a multi-type micro-grid cross-layer collaborative scheduling method based on multi-agent reinforcement learning, and relates to the technical field of power system and micro-grid scheduling, and the method comprises the steps: firstly constructing a multi-type micro-grid layered collaborative architecture, a multi-agent reinforcement learning scheduling model, and a centralized coordination layer where global agents are deployed in a platform layer; and the local agents are deployed at each distributed micro-grid node of the platform layer, a cross-layer cooperative training mode is adopted to train the model, and finally, cooperative decision results of the global agents and the local agents are integrated to generate a global-local cooperative scheduling scheme. According to the method, global optimization and local flexibility can be considered, and the hierarchical multi-agent architecture adopts a multi-type micro-grid hierarchical collaborative architecture, so that the method can be suitable for multi-type micro-grid operation scheduling of a novel power system, and multi-type micro-grid collaborative scheduling considering global optimization and local flexibility is realized.

The invention provides a data decoupling centralized query method and system based on MQ message triggering, and relates to the technical field of data query, and the method comprises the steps of receiving a message queue data change message, constructing a field, dividing check nodes by depending on a directed acyclic graph, carrying out data consistency verification, and pushing a transaction identifier to carry out concurrent operation scheduling. And collecting hotspot information by adopting a Bloom filter to adjust data fragmentation and cache distribution, generating a distributed query routing table, and distributing query requests according to the routing table. According to the method, the data consistency is improved, and the query performance is optimized.

The invention discloses a new energy low-carbon operation scheduling method and system for an island microgrid, and the method comprises the steps: generating a day-ahead scheduling plan with the minimization of the weighted average carbon emission intensity of a system as a core target through an upper-layer day-ahead optimization model based on load prediction data and renewable energy power prediction data; according to the day-ahead scheduling plan, generating an adjustment instruction of output of each unit through a lower-layer intra-day rolling optimization model; based on the adjustment instruction, calculating the wind and light abandoning power of the renewable energy source, and optimizing the renewable energy source consumption through a cost punishment mechanism; according to the optimized unit output and renewable energy consumption scheme, coordinating a charging and discharging strategy of an energy storage system to obtain an optimized scheduling instruction; and executing the scheduling instruction, and updating the system state through a closed-loop feedback mechanism. According to the embodiment of the invention, power balance and low-carbon operation of the island microgrid under a complex fluctuation working condition can be realized, and the utilization rate of renewable energy sources and the overall environmental protection property of the system are improved.

The invention discloses a multi-ship optimization method for cooperative operation and maintenance scheduling of an offshore wind plant, and belongs to the field of power system scheduling. According to the method, a heuristic algorithm and an accurate algorithm are combined, so that the problem of large-scale multi-ship collaborative maintenance scheduling of the offshore wind plant is solved; the method comprises the steps that a maintenance task constraint decomposition heuristic method is adopted to decompose a collaborative optimization scheduling problem into single-ship scheduling sub-problems, and it is ensured that constraint conditions of an original problem are met; performing path planning on each sub-problem by using an A * algorithm and a DATPF algorithm, and quickly and sequentially solving the sub-problems by using a CP-SAT solver; combining the solutions of single ship scheduling to form a comprehensive solution of a multi-ship collaborative maintenance scheduling problem; finally, a calculation result is obtained, and the effectiveness and robustness of the method are verified.

The invention discloses a network change operation scheduling method and device. The method comprises the steps that a conflict rule set and a to-be-executed network change operation set are acquired, and the conflict rule set is used for determining conflict reasons, conflict types and conflict eliminating rules between different network change operations; according to the conflict rule set, generating a conflict dependency graph of the to-be-executed network change operation set, the conflict dependency graph being used for representing a dependency relationship and a conflict relationship between operations in the to-be-executed network change operation set; and determining an execution sequence of all network change operations in the to-be-executed network change operation set according to the conflict dependency graph. The method provided by the invention at least solves the technical problem that the fault risk is relatively high after the network change operation is executed due to the fact that conflicts between the network change operations are not considered when the network change operation is executed in related technologies.

The invention provides a power distribution system area energy autonomous optimization method under new energy bearing capacity constraint. The method comprises the following steps: acquiring basic data of a target power distribution system area; a Voronoi-energy flow entropy dynamic space partitioning method is adopted, autonomous units are divided, and energy flow entropy features are calculated; generating initial new energy installed capacity, energy storage capacity and bearing capacity threshold values according to the space boundary and energy flow entropy characteristics of each autonomous unit; a Stackelberg game model is adopted, a dynamic electricity price, a carbon price and a deviation reward and punishment mechanism are introduced, a distributed iteration method is combined, and an operation scheduling strategy of each autonomous unit is determined; and performing rolling optimization and scheduling adjustment according to the feedback data, and dynamically updating the space boundary, the bearing capacity threshold and the scheduling strategy of the autonomous unit. According to the method, the energy management capability of the power distribution system is effectively improved, the new energy resource configuration is optimized, the scheduling cost is reduced, and the flexibility and stability of the system are enhanced.

The invention discloses a collaborative optimization configuration method and system for a light storage micro-grid in an expressway service area, and the method comprises the steps: firstly considering the photovoltaic bearing capability of the expressway service area, and constructing a master-slave game model of the light storage micro-grid; the master-slave game model comprises a capacity investment configuration master model with minimum investment cost and maximum photovoltaic bearing capacity as targets and an operation scheduling optimization slave model with minimum operation cost as targets; and solving the master-slave game model by nesting a commercial solver through adopting a Lu Eupe fox optimization algorithm, and obtaining a collaborative optimization scheme of capacity investment configuration and operation scheduling of the optical storage micro-grid in the expressway service area. According to the method, the master-slave game model of the optical storage micro-grid is constructed, the key problems of scattered resource allocation, insufficient operation coordination and the like of the optical storage micro-grid in the expressway service area are effectively solved, the power purchase cost of the power grid is effectively reduced, and the power supply reliability of the optical storage micro-grid is greatly improved.

An optimization scheduling method for coupling microgrid considering electro-thermal load demand coordination, comprising: building hydrogen-containing microgrid operation model and building electro-thermal load demand coordination response model by considering peak-valley complementary characteristics of original electric load and thermal load in system, characterizing electric load by time-of-use electricity price demand response method, by considering thermal load having heat transfer inertia and fuzziness of user temperature perception, adjusting electro-thermal load demand flexibly based on different electricity prices; based on hydrogen-containing microgrid operation model and electro-thermal load demand coordination response model, building objective function by minimizing sum of total operation cost, wind and light curtailed cost of renewable energy and demand response compensation cost, building electro-hydrogen coupling microgrid optimal scheduling model according to constraint conditions and solving it by mixed integer linear programming, to generate and send instructions based on solved optimal operation scheduling strategy to the microgrid for controlling corresponding unit equipment operation.

The invention discloses an intelligent ship lock system based on digital twinning, which belongs to the technical field of ship locks and comprises a data acquisition module, a data processing module, a model construction module, a simulation training module, an analysis and prediction module and a fault operation and maintenance module. The intelligent ship lock management and control system solves the problem that an existing ship lock system cannot monitor, manage, control and operate and maintain the ship lock in real time, so that the intelligent ship lock management and control effect is poor, ship lock operation scheduling automation, infrastructure maintenance management intelligence and navigation service information transparency can be achieved, the intelligent and integrated management level of the ship lock is effectively improved, and the intelligent ship lock management and control effect is improved. The ship lockage intelligent safety management and control is achieved, ship lock intelligent gear shifting and intelligent lockage guiding are achieved in the whole ship lockage process, the lockage safety level and the passing efficiency are comprehensively improved, the scientific decision making and comprehensive display ability is provided for a management department, and the intelligent ship lock management and control effect can be improved.

The invention relates to the technical field of supply chain management, in particular to a CBAM-based product full-life-cycle dynamic supply chain optimization method and management system, and the method comprises the steps: obtaining materials and numbers required by an outlet, building the association between raw materials and transportation links, collecting production carbon release details, and screening compliance paths to form an operation list. And judging the position overlap ratio to generate a resource path atlas, and outputting a supply chain scheduling scheme according to time consumption and span sorting. The method comprises the following steps: quantifying a geographic distance and a transportation mode in a raw material path, establishing mapping between transportation and a carbon source, identifying a carbon release path by combining an equipment operation section sequence, rejecting non-compliant paths according to target market limitation, associating a screening result with a carrying node space position, and constructing a connection combination according to a position overlap ratio. The path is optimized through time consumption and span sorting, and collaborative optimization of path screening and job scheduling is achieved.

The application discloses a power transmission line fault diagnosis and operation and maintenance scheduling method and system based on network learning, relates to the technical field of power transmission line fault diagnosis and operation and maintenance scheduling, and comprises the following steps: acquiring electrical quantity data, equipment operation state data, ring network topology and time sequence data and the like, extracting relevant data to construct a high-risk equipment area and a visual high-risk area heat map, combining the electrical quantity data to construct a visual risk classification map, combining a convolutional neural network-long short-term memory network hybrid model to construct an intelligent recognition storage network and a fault type classification recognition model, optimizing the model through network learning and an attention mechanism, combining an operation and maintenance management system to autonomously allocate a maintenance team and resources, and then realizing autonomous optimization and closed-loop operation, so that the application realizes full-process coverage from fault perception, intelligent decision and efficient response, significantly shortens the reaction time after line fault occurrence, and comprehensively enhances the fault disposal and operation and maintenance capability of the power grid system.

The invention provides a server delivery and operation and maintenance automation system oriented to multi-platform collaboration, and aims to solve the problems of split delivery process, lagging expansion response, high operation and maintenance complexity and the like in a heterogeneous infrastructure environment. According to the system, by constructing a multi-platform adaptive abstraction layer, bottom layer differences of private cloud, job scheduling platforms and the like are shielded, and unified modeling and standardization of delivery and operation and maintenance processes are achieved; on the basis, key links such as installation file generation, expansion file generation, configuration injection and deployment verification are integrated, an end-to-end one-key delivery assembly line is built, and rapid and consistent deployment of a cross-platform environment is supported; meanwhile, an automatic operation and maintenance closed loop is constructed, and operation and maintenance operations such as health examination, fault warning, configuration repair, service restart, node replacement and elastic capacity expansion and contraction are automatically executed by collecting operation states, logs and performance indexes of the server and the service in real time and combining a preset strategy or an intelligent analysis model. By means of the method, efficient delivery and intelligent capacity expansion of server resources are achieved, and the autonomous operation and maintenance capacity, the high availability and the service continuity guarantee level of the system in a complex multi-platform mixed deployment scene are remarkably improved.

Techniques for obtaining a Mobile Originating (MO) push schedule from a device in a wireless network are described herein. For example, the system may request MO push schedules from device endpoints and compare the MO push schedule with existing job schedules stored by the system. If no existing job schedule matches the MO push schedule, the system may generate a new job schedule and add the endpoint as a member of the job schedule. In some cases, if an existing job schedule matches the MO push scheduled, the system may add the endpoint as a member of the existing job schedule. In some examples, the system may receive an indication that the MO push schedule of the endpoint has changed and may add and / or remove the endpoint from a job schedule based on an updated MO schedule received from the device endpoint.

The invention provides a dynamic predictive job scheduling method, which comprises the following steps of: acquiring a configuration file of a to-be-scheduled EDA (Electronic Design Automation) task by a scheduling central system, and extracting job characteristics of the to-be-scheduled EDA task; acquiring real-time monitoring data of the current running task from the computing node cluster; the prediction engine predicts the to-be-scheduled EDA task and the current running task according to the job characteristics and the real-time monitoring data to obtain first prediction data of the to-be-scheduled EDA task and second prediction data of the current running task, and the first prediction data and the second prediction data are fed back to the scheduling center system; the scheduling center system sends the first prediction data, the second prediction data and the real-time monitoring data to a decision engine; the decision engine determines a resource preemption strategy of the to-be-scheduled EDA task and feeds back the resource preemption strategy to the scheduling center system; the scheduling center system determines the execution sequence of the EDA tasks to be scheduled and sends the execution sequence to an execution engine; and the execution engine performs resource allocation on the to-be-scheduled EDA task.

The invention discloses an integrated energy system multi-body operation scheduling method and system based on a mixed game, and the method comprises the steps: building a multi-energy cooperation system architecture based on the characteristics and energy conversion relation of each energy body in an integrated energy system, and building each energy body model; based on each energy main body model, designing a three-layer mixed game model which comprises an upper-layer master-slave game model, a middle-layer cooperative game model and a lower-layer non-cooperative game model; the three-layer mixed game model forms closed-loop linkage through information interaction between layers; and solving the three-layer mixed game model, performing benefit distribution and settlement on each energy main body based on a solving result, and realizing operation scheduling of each energy main body. According to the method, the problem that a traditional single game framework cannot cover multi-level requirements is effectively solved, and the multi-agent collaboration efficiency and the market activeness are remarkably improved.

The invention discloses a gate distributed operation scheduling method and system. According to the method provided by the invention, a'global-regional-device 'three-level hierarchical collaborative architecture is constructed, and the overall operation condition of the drainage basin is identified and the global scheduling instruction with maximized revenue is generated through the global optimization server based on the reinforcement learning model under the driving of the dynamic reward function; and the achievement of high-level targets such as water resource utilization efficiency or flood control safety is ensured. And the cooperative control gateway responds to the global scheduling instruction so as to carry out data fusion and cooperative analysis on data in the region through a supply and demand balance model, and the global instruction is converted into a cooperative scheduling strategy adaptive to the region characteristics. And finally, the control device accurately executes the localized instruction to form a sensing-decision-execution closed loop. And self-learning and self-optimization of the scheduling strategy are realized, and gate number expansion and scheduling target dynamic adjustment are supported.

The invention provides a double-layer optimization scheduling method for source network load storage collaborative loss reduction, and belongs to the technical field of source network load storage collaborative scheduling. A double-layer optimization framework comprising an upper layer planning model and a lower layer operation model is constructed, and a key node set is screened by using a node importance comprehensive scoring method based on a graph theory; the improved wolf pack algorithm is adopted to solve a double-layer coupling problem, a lower-layer operation effect is fed back to an upper layer to serve as a fitness evaluation basis, and in cooperation with population aggregation degree monitoring, a wandering wolf random jumping mechanism and a reverse learning strategy, deep collaboration and global optimal solution of a planning layer and an operation layer are achieved. The technical problem of poor network loss optimization effect caused by lack of an effective double-layer coupling solution mechanism for collaborative optimization planning and operation scheduling of the distributed photovoltaic and energy storage system in the power distribution network is solved.

The invention discloses a self-adaptive micro-grid operation optimization method, system and device and a medium, and belongs to the technical field of micro-grid operation optimization, and the method comprises the steps: obtaining the operation state data of a micro-grid; performing depth feature coding and space-time attention processing on the micro-grid operation state data through a state sensing module to generate unified state representation; carrying out online strategy learning and optimization through a strategy evolution module based on unified state representation, and outputting a strategy optimization result; a new energy uncertainty model is adopted to generate multi-scene prediction data, the prediction data and a strategy optimization result are combined, and a prediction control problem is solved through a rolling optimization module; and according to a real-time risk assessment result, the decision weights of the strategy evolution module and the rolling optimization module are adjusted, and a micro-grid operation scheduling instruction is output. According to the method, a new state deep perception-dynamic strategy learning-robustness optimization normal form is constructed, and intelligentization and adaptive learning of the micro-grid are realized.

The invention discloses a multi-working-condition collaborative operation scheduling method for a wheel bucket excavator, a medium and computer equipment. The method comprises the following steps: acquiring operation data of the wheel bucket excavator and transportation equipment in real time, wherein the operation data comprises an operation equipment state, an operation task emergency degree and a material value; and extracting working condition characteristics such as excavation strength of the wheel bucket excavator under stripping and mining working conditions. An evaluation matrix is constructed by using an improved analytic hierarchy process, a digital twinning model is constructed in combination with data and features, and online learning is performed by using a federated learning framework. And dynamically adjusting task priorities by utilizing learning results, calculating a matching degree, determining a scheduling strategy and scheduling equipment until all tasks are completed. By constructing a multi-source sensing and digital twinborn model and combining a dynamic priority evaluation and self-adaptive task allocation technology, intelligent collaborative scheduling of the wheel bucket excavator under multiple working conditions is realized, so that the working efficiency, the equipment utilization rate and the anti-disturbance capability are improved, and the purposes of reducing energy consumption and guaranteeing production continuity are achieved.

The invention discloses a locating and sizing method and device for an intelligent soft switch in a power distribution area, terminal equipment and a storage medium, and belongs to the field of power distribution network optimization. The method comprises the following steps: constructing a bilevel planning model, a power distribution network operation constraint and an energy storage equipment constraint according to electrical parameters, distributed energy parameters, energy storage parameters and cost data; by taking the minimum total operation cost of the system as a target, repeatedly executing a locating and sizing optimization operation on the bilevel programming model until a preset number of iterations is reached, and obtaining the installation position and the installation capacity of the intelligent soft switch; the optimization operation comprises solving the current installation position and capacity of an upper-layer intelligent optimization model of the bilevel planning model; if the preset number of iterations is reached, the final installation position and capacity are determined; and if not, inputting into a lower-layer operation scheduling model to construct intelligent soft switch operation constraints, solving the lower-layer model under the three constraints, and updating the total operation cost of the current system. According to the invention, the problem of multi-zone-area multi-intelligent soft switch collaborative optimization in the prior art is solved.

The invention relates to the technical field of computers, and discloses a tail delay optimization job scheduling method and system based on a heterogeneous GPU cluster. The method comprises the following steps: constructing a cluster physical topological graph containing link delay, bandwidth and hop count; predicting a job communication demand graph through static code analysis and a graph neural network; executing topology matching scheduling, and preferentially deploying a high communication task pair on a high-speed interconnection link in a node or a low-delay link on the same rack; and tail delay is monitored during operation, and cost-aware local rescheduling is triggered. The system comprises a physical topology modeling module, a communication demand prediction module, an affinity scheduling module and a dynamic rescheduling module. Through active prevention and closed-loop optimization, job tail delay is reduced, and heterogeneous cluster service quality and resource utilization efficiency are improved.

The invention relates to the technical field of port container loading and unloading, in particular to a system and method for container lock disassembly and assembly operation scheduling, and the system comprises a quay crane terminal subsystem which is arranged in a quay crane cab and is used for transmitting locking instruction information to a server through a wireless network; the server is used as an information center for receiving and processing the locking instruction information sent by the quay crane terminal subsystem and pushing the locking instruction information to the corresponding container truck terminal subsystem for transporting the container; the container truck terminal subsystem is arranged in a container truck cab and used for receiving and displaying the locking instruction issued by the server, and a terminal screen can indicate whether the currently transported container needs to be locked or not; and the LED indication screen is connected with the container truck terminal subsystem through a network cable and communicates with the container truck terminal subsystem through a binary code, and the problems that in the prior art, efficiency is low, the error rate is high, and the process is not traceable are solved.

The invention relates to a GPU equipment isolating and mounting method and device, equipment and a storage medium, and relates to the field of artificial intelligence chips. The method comprises the following steps: CPU equipment in computer equipment acquires a GPU resource request submitted by a user through a job scheduling system, determines target GPU equipment from candidate GPU equipment based on the GPU resource request so as to allocate the target GPU equipment to the user, and assigns the target GPU equipment to a visible environment variable defined by a manufacturer of the target GPU equipment; and calling the user space container system through a container arrangement tool of the job scheduling system, so that the user space container system creates a container according to the visible environment variable, analyzes the visible environment variable to obtain target GPU equipment, and mounts the target GPU equipment into the container. By adopting the method, the flexibility and efficiency of GPU resource management can be improved.

The invention relates to the technical field of power systems, in particular to a graph attention neural network-based power distribution network topology identification method, system and device and a medium, and the method comprises the steps: collecting the voltage, active power, reactive power and other operation state quantities of a power distribution network node, and constructing a topology identification sample; performing standardization processing on the collected data, and unifying node features; after standardizing the data, generating an adversarial sample by adopting an adversarial strategy, and taking the adversarial sample and the topology identification sample as an input training network model; acquiring node data in real time, inputting the node data into the trained graph neural network model, and outputting a topological state result; through combination of sample construction with topological labels, data standardization processing, adversarial sample generation and model training, real-time reasoning output is carried out. According to the method, a complete processing flow is formed, limitation caused by dependence on single simulation data and measurement difference is solved, it is guaranteed that the topological state of the power distribution network can still be recognized under complex working conditions, and a result capable of being used for operation scheduling and analysis is output.

The application relates to the field of sanitation operation scheduling, in particular to a sanitation operation scheduling method based on the Internet of Things, which comprises the following steps: determining an iterative stop condition of a load CEEMDAN model based on time local weighted variance of load data, so as to generate vehicle load characteristics; generating vehicle oil consumption characteristics through an oil consumption CEEMDAN model based on statistical values of load data residuals and oil consumption data residuals, so as to calculate oil consumption channel residual signals; generating comprehensive vehicle carrying characteristics through a feature extraction model based on a multilayer perception machine, a gating mechanism and a cross-modal attention mechanism; and generating corrected vehicle garbage load through a long short-term memory network and a fully connected layer architecture. The application realizes efficient noise reduction and accurate decomposition of non-stationary time series data of oil consumption and load, realizes accurate processing of Internet of Things data of garbage transfer vehicles, improves the operation efficiency of the garbage transfer vehicles and reduces the operation cost.

The invention discloses a wind power prediction covariable offset adaptive method based on a conditional energy model. According to the method, the wind power generation power prediction model is constructed by utilizing the gated cycle unit network, and offline training of the wind power generation power prediction model is completed by adopting a sample weighting mechanism, so that the robustness of the wind power generation power prediction model to distribution change is enhanced. A conditional de-noising score matching strategy is adopted to learn the distribution difference of data in a training stage and a prediction stage through a conditional energy model, and a sample weight used for measuring the covariable offset degree is obtained based on the model. Data samples flowing in real time are stored in a replay buffer area, incremental learning is carried out on a condition energy model and a wind power generation prediction model through an online updating mechanism, and the prediction performance is kept stable. The method can effectively improve the power prediction precision and operation scheduling capability of the wind power plant under complex meteorological conditions, and has good engineering practical value and deployment flexibility.

The invention provides a construction engineering machinery collaborative operation scheduling method based on a cloud platform, and the method comprises the steps: carrying out the multi-dimensional division of an operation range of a mechanical device through employing a grouping analysis algorithm according to a real-time position data set, and determining a dynamic operation coverage region reflecting the motion characteristics of the device; for the mark set, obtaining time sequence motion information of an adjacent mechanical device, simulating an adjusted device moving path through a trajectory optimization algorithm, and obtaining an optimized path set for avoiding an initial conflict; according to the corrected posture set, generating a control instruction set for position adjustment, distributing the control instruction set to each mechanical device terminal through the cloud management platform, and obtaining feedback confirmation information after device execution; and verifying whether the spatial conflict is eliminated from the feedback confirmation information to obtain a final scheduling scheme set integrating all adjusted schemes.

The present application belongs to the field of medical information technology and intelligent insurance pricing, and relates to a disease-specific risk financing optimization method and device based on a double-layer value feedback loop, which comprises the following steps: collecting real-time multi-modal clinical data of patients, extracting time series derivatives, embedding semantics and fusing features to construct a global benchmark feature tensor; inputting a deep neural network to derive a high-dimensional clinical risk vector through cross-attention; calculating the actual settlement amount through logarithmic probability transformation, hyperbolic tangent boundary mapping and historical load negative feedback; packaging as a double-domain encrypted certificate and binding with the inpatient medical record number atomization to realize event-level precise mapping; triggering model sandbox retraining and hot loading with residual variance to form the first feedback loop of clinical counter-benefiting; reading the economic relief measurement value to dynamically adjust the message queue, cache preloading and container computing power to form the second feedback loop of operation scheduling. Cross-subsidies can be eliminated, model self-evolution and mathematical-level risk control can be realized, and the method is suitable for commercial health insurance and other scenarios.