239 results about "Decision rule" patented technology

The invention provides a crop growth state evaluation method based on the agricultural Internet of Things, and the method is characterized in that the method specifically comprises the following steps: S1, collecting the multi-dimensional data of a crop growth environment in real time, and outputting an original data set; s2, aligning the multi-dimensional data according to geographic coordinates, extracting features, and constructing a multi-modal fusion feature vector; s3, automatically identifying the current growth stage of the crop and the confidence of the current growth stage according to the fused feature vector; s4, calculating a crop health degree score and grading according to the growth stage information and the feature data, and identifying a stress state detection result at the same time; s5, according to a quantitative evaluation result, predicting a growth trend and generating a suggested decision scheme; and S6, the decision is monitored, the decision execution effect is fed back to the system, and the model parameters and the decision rules are continuously optimized. A closed-loop mechanism of evaluation, decision, feedback and optimization is formed, the applicability and evaluation precision of the system are continuously improved, and agricultural production is promoted to be upgraded to precision and intelligence.

The invention belongs to the technical field of flight simulator maintenance, particularly relates to a flight simulator predictive maintenance method based on machine learning, and solves the problems that existing maintenance depends on regular inspection and passive maintenance, fault early warning lags behind, and the false and missing report rate is high. The method comprises the following steps: acquiring historical operation data, sensor time sequence data, fault records and environmental parameters of a flight simulator, and carrying out cleaning, labeling and feature fusion preprocessing on the historical operation data, the sensor time sequence data, the fault records and the environmental parameters; constructing a composite health feature set containing statistical features, dynamic health state values and aerial material reliability parameters; a mixed prediction model (random forest feature screening + LSTM time sequence prediction + adaptive correction reliability evaluation) is adopted to train a model, prediction result fusion analysis and multistage decision rule post-processing are combined, and a maintenance work order and a spare part demand plan are output. According to the method, the accuracy and timeliness of fault prediction are improved, the maintenance conversion from passive response to active pre-judgment is realized, and the maintenance cost and the non-planned shutdown risk are greatly reduced.

The invention relates to the technical field of tunnel engineering automation and intelligent control, and discloses a TBM tunneling parameter intelligent optimization decision-making system based on an LSTM network, and the system comprises a data collection and preprocessing module which obtains external data and outputs a tunneling parameter sequence; the probabilistic tunneling trend prediction module is used for outputting a prediction expected value and prediction uncertainty; the prospective geological precursor sensing module is used for matching and identifying known risks and outputting alarm events; and determining an optimal tunneling mode by the dynamic risk avoidance decision matrix. When the prediction uncertainty is too high, activating the prospective template driven by the uncertainty to excavate a new precursor template and update the template library; meanwhile, the decision-efficiency relevance evaluation and strategy self-optimization engine optimizes the decision rule according to the actual tunneling efficiency. According to the method, decision making is carried out through quantitative risk prediction and fusion of multi-source information, and a double learning closed loop of knowledge discovery and strategy optimization is established, so that the reliability, the adaptability and the long-term efficiency of system decision making are remarkably improved.

The invention discloses an elderly cognitive function adaptive training method and system based on multi-modal interaction feedback, and relates to the technical field of smart medical treatment. The method comprises the steps that basic information of a user is collected for initial cognitive ability evaluation, a user cognitive portrait is constructed according to an evaluation result, and an initial training task with the corresponding difficulty is allocated; collecting multi-modal interaction data in real time according to the initial training task; carrying out fusion analysis on the multi-modal interaction data by utilizing a machine learning model to obtain a quantized real-time state index; based on the real-time state index and the performance data of the current task, dynamically adjusting a subsequent training task through an adaptive decision rule engine; all-dimensional data of each training task is recorded, a visual cognitive competence development trend report is generated through longitudinal comparative analysis, and a machine learning model and a self-adaptive decision rule engine are continuously optimized and trained by utilizing accumulated user data to form an optimized training closed loop. The cognitive function training effect of the old people can be improved.

The invention provides a greenhouse water, fertilizer and pesticide optimization control system and method based on environment-crop-input multi-element coupling, and the method comprises the steps: deploying a sensor network, collecting greenhouse data, processing the greenhouse data, generating a space-time continuous data set, and carrying out the optimization control of the greenhouse water, fertilizer and pesticide on the basis of the generated space-time continuous data set. And quantifying a dynamic causal chain of the environment, the crops and the input, generating a decision rule base, fusing the decision rule base with a real-time simulation result of the three-dimensional digital twinborn model of the root zone, and generating a water, pesticide and fertilizer regulation and control instruction. Space-time continuous acquisition and fusion processing of greenhouse environment-crop-input product data are realized through a multi-modal sensor network, a decision rule base containing a dynamic causal chain is constructed, real-time simulation of a root zone three-dimensional digital twinborn model and Bayesian causal reasoning are combined, an accurate water, fertilizer and pesticide regulation and control instruction is generated, and the accuracy of the water, fertilizer and pesticide regulation and control is improved. The nonlinear causal relationship among multiple elements is effectively quantified, and the decision scientificity is improved.

The invention discloses an emergency pre-examination grading system and method based on collaborative decision of a large language model and a tree model. The method comprises the following steps: constructing a data set according to patient information and screening data; training an initial random forest model based on the data set, generating a basic decision rule, mining and extracting high-frequency features based on association rules, and combining to obtain a candidate rule pool; constructing a cue word structure adaptive to the field, driving LLM to complete rule correction, and forming a correction rule set; performing multiple rounds of rule random division and rule combination generation based on the correction rule set, and then screening out an optimal rule; and on the basis of a specific scene, the expert rule and the corrected optimal rule are fused, matching is carried out on patients, and emergency pre-examination grading is realized. According to the method, the interpretability is improved while the grading accuracy is improved, and the problem of poor cross-courtyard generalization of a machine learning model is effectively solved. The generalization ability of the rule is remarkably improved, and the method is adaptive to a multi-center combined diagnosis and treatment scene.

The invention relates to the field of production line state monitoring, and particularly provides a production line state monitoring and MES integration notification method and system, and the method comprises the steps: collecting multi-modal data, and constructing a unified state context object; performing space-time alignment and dynamic weight fusion based on a process stage on the multi-source asynchronous data to obtain a feature vector; reasoning a semantic current state of the production line through a time sequence mode recognition model; generating an enhanced data set based on the state context and the current state of the production line, and outputting a structured state event through anomaly detection and root cause analysis; outputting an optimal response strategy through a predefined decision rule in combination with MES service information; and packaging the strategy into an operation instruction which can be identified by the MES and issuing the operation instruction, and driving the MES to execute work order creation, billboard updating or message notification. According to the method, semantic diagnosis, global decision and automatic response of the production line state are realized, so that the production transparency, the decision accuracy and the operation and maintenance automation level are improved.

The invention discloses a low-speed unmanned vehicle artificial intelligence decision and performance evaluation system, and belongs to the technical field of low-speed unmanned vehicle scheduling, and the system comprises a generation module which is used for generating a scene novelty index through monitoring the instantaneous offset of an environment parameter; the topology module is used for constructing a dynamic scene topology according to the scene novelty index, and the dynamic scene topology comprises nodes and edge weights generated based on the scene novelty index; the decision-making module is used for converting the edge weight into a decision-making rule according to the dynamic scene topology and generating a decision-making strategy; the performance calculation module is used for tracking the deviation between the vehicle motion parameter and the control instruction during the execution of the decision strategy to calculate a performance disturbance coefficient; and a topology calibration module. The method can effectively adapt to unrecorded edge scenes such as sudden congestion and severe weather superposition, improves the generalization ability of a system to complex scenes exceeding a pre-training range, and reduces decision errors caused by insufficient scene adaptation.

The invention provides a flange assembly predictive maintenance method based on residual life distribution dynamic identification, which comprises the following steps: firstly, establishing a linear Wiener model of a flange assembly degradation process, and designing a Bayesian parameter dynamic updating mechanism based on normal-inverse gamma conjugate prior; secondly, deducing residual life complete probability distribution considering parameter uncertainty through a Monte Carlo sampling method, overcoming the limitation of point prediction, and providing a maintenance decision rule based on a time-probability threshold; and finally, establishing a decision parameter optimization model with the goal of minimizing the long-term average cost rate, and solving an optimal maintenance strategy through system simulation. Compared with the prior art, the residual life prediction accuracy is remarkably improved, the maintenance cost and the equipment reliability are effectively balanced through a probabilistic decision-making mechanism, the full-life-cycle maintenance cost is remarkably reduced while the flange sealing safety is ensured, and the method has important popularization value in engineering equipment predictive maintenance.

The invention provides a cloud intelligent gas leakage analysis and decision-making method based on lightweight discrimination, and the method comprises the steps: enabling a mobile inspection platform to movably collect the gas leakage monitoring information of each monitoring point in a monitoring region, and determining a lightweight discrimination index; the mobile inspection platform determines a hard threshold discrimination result by adopting a hard threshold decision rule based on the lightweight discrimination index; if it is judged that the risk level is high, local response is immediately carried out, and a preset decision scheme is matched; if the risk level is judged to be low, edge node region decision is started; an edge node calculates a gas concentration accumulated value, a concentration fluctuation intensity degree and a sudden increase leakage amplification early warning factor, the gas concentration accumulated value, the concentration fluctuation intensity degree and the sudden increase leakage amplification early warning factor are input into a LightGBM time sequence classification model together with a lightweight discrimination index, a leakage risk level is determined according to model output and serves as a soft threshold discrimination result, and a preset decision scheme is matched; and the cloud model optimizes and collects conflict samples to carry out incremental training on the model of the edge node. According to the invention, the missed alarm rate can be reduced, and the discrimination robustness is improved.

The invention discloses a heterogeneity causal effect quantitative evaluation system and method for electric vehicle charging behaviors, and relates to the field of electric vehicle big data analysis, artificial intelligence and causal inference. The system comprises a data preparation and variable system construction module, a heterogeneity causal effect quantitative evaluation model construction module and a model application and result interpretation module. A dual machine learning framework is adopted, systematic deviation of high-dimensional hybrid factors is eliminated by calculating orthogonalization residual errors, a causal forest model is constructed on this basis, a conditional average processing effect is accurately and stably quantified based on a splitting criterion of maximizing effect heterogeneity, and a high-precision and high-precision effect is obtained. The problem of causal effect estimation deviation of a traditional method under the forms of high-dimensional data and complex functions is solved. In order to enhance the interpretability, an agent model based on a single decision tree is further constructed, and a complex causal forest conclusion is extracted into a group of visual and operable If-Then decision rules, so that a complete closed loop from data to robust and interpretable decisions is realized.

Various embodiments include a system that utilizes machine learning to detect and mitigate bot attacks. The system comprises processing circuitry. The processing circuitry obtains historical traffic data and attack traffic data in response to an attack notification. The attack traffic data characterizes traffic received during a bot attack and the historical traffic data characterizes other traffic received when the bot attack is not occurring. The processing circuitry extracts features from the historical traffic data and the attack traffic data. The processing circuitry trains a machine learning classifier to identify the features that correspond to attack traffic and the features that correspond to legitimate traffic. The processing circuitry forms decision rules based on an output from the machine learning classifier to block the attack traffic based on the features that correspond to the attack traffic. The processing circuitry generates one or more security policies based on the decision rules.

The invention relates to the technical field of data processing, in particular to a power grid probability equivalent model construction method and device based on multi-scene clustering, and the method comprises the steps: obtaining a power grid multi-dimensional data source, executing a spatial-temporal feature decoupling operation, and outputting a spatial-temporal feature tensor with a weight identifier; executing graph-driven enhanced clustering operation based on the tensor, constructing a topological risk graph structure, and generating a safety margin offset thermodynamic diagram through a deep reinforcement learning decision feature compression strategy; executing a double-algorithm optimization traceability operation, adopting a genetic algorithm to encode a feature cutting operation, taking a margin deviation value as a fitness function, optimizing a time window parameter by a butterfly algorithm, and outputting a feature operation traceability matrix; and feeding back the matrix to the enhanced clustering operation to update the decision rule. According to the method, the problem of dynamic security risk quantification distortion caused by scene clustering dimension reduction is solved under the constraint of limited computing resources through risk perception feature processing, enhanced resource allocation and a closed-loop architecture of double-algorithm traceability optimization.

The invention relates to the technical field of forklift control, in particular to an efficient and energy-saving off-road forklift electric drive system and a control method. Multi-source data is collected through a multi-source sensor, and data preprocessing is carried out; outputting a control decision instruction according to a priority decision rule based on the preprocessed data in combination with a fuzzy logic algorithm; in combination with the working condition characteristics and the real-time data, dynamic parameter adjustment is performed on the rotating speed, the torque and the current parameters of the motor by utilizing a reinforcement learning framework, and a PWM duty ratio control instruction is generated; on the basis of a PWM duty ratio control instruction and a hydraulic pressure signal fed back by a hydraulic system, the same-frequency anti-phase compensation torque is calculated, a signal is injected to suppress fluctuation, and torque distribution and hydraulic flow are adjusted through a ramp function; and establishing a hierarchical health early warning mechanism according to the historical data, judging the current early warning level, and outputting a maintenance strategy instruction according to the early warning level. According to the scheme, equipment health management and stability are enhanced through reinforcement learning dynamic optimization and graded early warning.

The invention relates to the technical field of environment control, in particular to a cold and heat source environment control method and device with an air purification function, and the method comprises the steps: collecting various parameters through a sensor network, and carrying out the data fusion based on a decision rule base to select a dominant working mode; corresponding air treatment operation is executed, and a composite energy-saving strategy including precooling optimization and phase change latent heat assistance is triggered under the refrigeration working condition; fuzzy control and machine learning are applied to optimize system output and predict load, and a decision rule base and algorithm parameters are updated through self-learning; air supply and exhaust power is controlled to maintain the negative pressure gradient of the key area, and an air path distribution mechanism is adjusted to organize a one-way air flow line; the device comprises an intelligent control module, a sensor group, a fresh air module, a purification module, a cold and heat source module, a bypass pipeline and an air path distribution mechanism. The problems that an existing system is poor in function collaboration, low in energy efficiency and unreasonable in airflow organization are solved, and intelligent decision making, deep energy saving and accurate environment control are achieved.

The invention relates to the technical field of data monitoring, in particular to a workshop operation state real-time monitoring, management and control method and system, and the method comprises the steps: carrying out the real-time collection of multi-source heterogeneous data of each process execution unit through a plurality of data collection terminals, and carrying out the standardization processing of the collected multi-source heterogeneous data; based on the standardized multi-source heterogeneous data, extracting multi-dimensional features corresponding to each process execution unit, associating the multi-dimensional features with plan standard features, inputting the associated features into a state judgment model, and outputting a comprehensive state identifier; if it is judged that the comprehensive state identifier triggers a state abnormal event, comparing the state abnormal event with a preset decision rule base, and outputting corresponding decision action information; and packaging the decision action information into a corresponding management and control instruction, and transmitting the management and control instruction to the corresponding process execution unit for execution through an instruction distribution interface, thereby realizing comprehensive, accurate, real-time and intelligent management and control of the workshop operation state.

The invention discloses a power supply modular design method and system, and relates to the technical field of computer-aided process design, and the method comprises the steps: receiving power supply design parameters, carrying out the feature coding through a three-layer neural network, generating an initial design vector, and constructing an SNN decision rule base according to the initial design vector; performing parameter evolution on the SNN decision rule base through a pulse frequency mapping function and a dynamic current adjustment function to generate a collaborative decision packet, and injecting a noise suppression factor into the collaborative decision packet to generate a noise immune decision packet; and constructing an electromagnetic field distribution model, inputting the noise immune decision packet into the electromagnetic field distribution model to calculate the electromagnetic coupling strength of the pulse signal, and comparing the electromagnetic coupling strength with a preset strength threshold to generate an electromagnetic isolation enhancement decision packet. According to the invention, through neural network feature coding and a full-closed-loop parameter transmission mechanism, lossless conversion and dynamic response from power supply design parameters to manufacturing instructions can be realized.

The invention discloses a radar radiation source identification method based on zero sample learning, a storage medium and equipment, and the method comprises the steps: firstly obtaining an effective pulse of a radar radiation source signal, and training a one-dimensional convolutional neural network containing an auto-encoder based on the effective pulse of a known signal; the method comprises the following steps: selecting an optimal discrimination channel which is most effective for category discrimination by analyzing initial feature space distribution, determining a sparse region of known category signal distribution along the channel as a rejection interval, and identifying an unknown signal; then, unknown category signals are combined, and an enhanced open set classifier capable of explicitly distinguishing unknown categories is trained; and finally, integrating the initial model and the enhanced open set classifier, and identifying an unknown category and classifying a known category through a confidence-based fusion decision rule. According to the method, the detection capability of the unknown radar radiation source can be learned from known data without a prior sample of the unknown radar radiation source, and the identification problem of the unknown radar radiation source in the open environment is effectively solved.

The invention relates to the technical field of people flow simulation analysis, and particularly provides a people flow simulation analysis method and system based on spatio-temporal behavior dynamics, and the method comprises the steps: obtaining urban multi-dimensional spatio-temporal heterogeneous data, and carrying out the standardization; constructing a dynamic coupling mechanism of crowd behaviors, spatial constraints and industrial function supply, and extracting travel decision rules and destination selection logic; by taking the mechanism as a constraint, constructing a space-time behavior dynamic model in combination with subject modeling, simulating group flow space-time evolution and outputting crowd flow characteristic parameters; and based on the parameters, predicting the people flow distribution and aggregation dissipation trend of the target scene, and performing visual display. According to the method, through combination of multi-factor dynamic coupling and subject modeling, simulation accuracy and scene adaptability are improved, a prediction result is visual, and scientific decision support can be provided for urban planning, traffic management, public safety guarantee and the like.

The invention provides a DIKWP semantic modeling method. Complex problems of an enterprise are converted into a five-layer linkage semantic map. The method comprises the following steps: 1, cleaning multi-source heterogeneous data to form nodes by a data layer; (2) information layer construction domain ontology clarification concepts and constraints; (3) the knowledge layer aligns with the knowledge base to infer and complement the relationship to generate a knowledge graph; (4) the wisdom layer outputs scheme elements, benefits and risks based on decision rules; and (5) modeling each intention layer refining strategic target constraint layer. Full-link semantic mapping from data to intention is achieved, consistency and accuracy of problem definition are improved, cross-department implicit causality is mined, it is ensured that a scheme is accurately aligned with a strategic target, an interpretable semantic basis is provided for automatic scheme generation, scene simulation and optimization decision, and the method has remarkable commercial application value.

The invention relates to the technical field of fan blade life prediction, and discloses a large fan blade life prediction method based on a simulation model. The method comprises the following steps: acquiring a stress-strain data stream and an operating environment data stream of a blade, and analyzing the stress-strain data stream by means of a preset life constraint index to obtain a blade life distribution characteristic; combining a prediction decision rule base and a life fitness condition, integrating the life distribution characteristics and the operation environment data flow to generate a preliminary life prediction scheme, and pre-judging the leaf life distribution characteristics and environment data change by adopting a trend analysis technology to obtain predicted life distribution characteristics and predicted environment data; and optimizing and correcting the initial scheme according to the prediction data to form an optimized life prediction scheme, thereby realizing the life prediction of the large-scale fan blade. According to the method, multi-source data are integrated, trend analysis and scheme optimization are combined, prediction reasonability and accuracy are improved, and reference is provided for blade operation and maintenance.

The invention provides an equipment predictive maintenance method and system based on residual life quantiles, and belongs to the field of equipment maintenance and reliability engineering. According to the strategy, firstly, a Gamma process is adopted to construct an equipment degradation model, and Beta distribution is adopted to establish a residual degradation amount model; then designing a discrete check strategy based on the residual life quantile, determining a calculation mode of a discrete check interval, and formulating a maintenance decision rule; calculating maintenance related cost, including preventive maintenance cost and operation cost; and finally, constructing an optimization model with the goal of minimizing the average cost rate, and determining optimal maintenance parameters by adopting a discrete approximate iteration method. The method can accurately grasp the state change of the equipment, formulate a scientific and reasonable maintenance strategy, effectively reduce the maintenance cost, improve the reliability of the equipment, and is suitable for predictive maintenance of various industrial equipment.

The invention discloses a poplar growth prediction and region optimization method based on multi-source data fusion, and the method comprises the steps: collecting and preprocessing original multi-source heterogeneous data, generating the multi-source heterogeneous data, carrying out the data fusion processing, and generating a multi-source fusion data set; evaluating the influence weight of each variable in the multi-source fusion data set on the growth index, determining the driving factor type and the interaction of each variable, and constructing a growth model system and a single-variable simplified growth model based on the driving factor sorting and the interaction; generating a dynamic query data table, an economic mature age judgment result, a carbon sink calculation result and a thinning suggestion according to the constructed model; and based on the residual sequence of the long-period prediction model and the time sequence of the climate-driven factor, detecting the period coupling relationship between growth and climate, verifying the lock correlation of the climate-growth system, and determining a climate adaptability operation decision rule based on the period coupling relationship so as to generate a standardized operation scheme. According to the invention, the accuracy of growth prediction and regional optimization can be improved.

The invention provides an information system elastic control method for coping with sensor and actuator combined network attacks, which comprises the following steps of: constructing a tag Petri network model of an information system for describing a system structure and observation information, and depicting a control specification by adopting a generalized mutual exclusion constraint; in the offline analysis stage, a network attack model is constructed and used for analyzing the influence of network attacks on an information system; based on a system structure, designing a judgment rule for controlling the satisfiability of the specification under the network attack, and designing a decision rule; and an online control stage: collecting an observation sequence generated by a sensor channel in a running process of the information system in real time, judging whether a current system state meets a control specification or not, and controlling the information system according to a decision rule. According to the method, on the premise of not depending on system reachability analysis and state estimation, it is ensured that control specifications are still met in a combined attack scene, and the operation safety of the system is improved.

The invention discloses a power transaction auxiliary decision processing method and system based on extreme weather, and relates to the technical field of power transaction decision, and the method comprises the steps: obtaining an extreme weather early warning message to form an original weather early warning data set; performing space-time reference alignment on the data set to generate a grid chart containing a meteorological intensity matrix and a power generation facility distribution index; mapping the meteorological intensity matrix to obtain a meteorological sensitivity coefficient matrix, and carrying out topological association on power generation facility distribution indexes to obtain a power generation side asset vulnerability association table; inputting the two into a risk conduction calculation engine for coupling deduction to generate a power supply-demand imbalance risk thermodynamic diagram; constructing a transaction price elasticity prediction model based on the thermodynamic diagram, and forming a transaction strategy reference matrix; quotation interval suggestions are generated through matching of the decision rule base and are converted into standardized declaration instruction streams to be pushed to the transaction interface end. According to the method, the extreme weather information and the power transaction decision can be accurately joined, and the accuracy and the normalization of the transaction decision are improved.

The invention discloses an osmotic pressure analysis method and device fusing model optimization and physical inversion and a medium, and the method comprises the steps: calculating correlation coefficients of a specified reservoir water level and osmotic pressure under different time lags based on cross-correlation function analysis, and recognizing a differential nonlinear feature corresponding to the maximum correlation coefficient; aiming at the differentiated nonlinear characteristics of different dam types, constructing a plurality of corresponding diagnosis prediction models in parallel, and establishing a model selection decision rule to select an optimal diagnosis model; according to delay parameters obtained through cross-correlation analysis, a permeability coefficient is reversely deduced through a pore medium heat transfer diffusion theoretical formula; according to the regression slope and the theoretical value attenuation factor of the optimal model, a permeability coefficient is reversely deduced in combination with a correction diffusion equation, credibility evaluation and weighted fusion are performed on the permeability coefficient, and a comprehensive permeability coefficient is calculated; and diagnosing the soil type and the permeability characteristic grade. According to the method, the propagation delay characteristic of reservoir water level change in the pore medium can be accurately quantified, and the inversion precision of the permeability coefficient and the engineering applicability are improved.

The invention discloses a multi-dimensional network intrusion behavior intelligent identification method based on deep learning. The method comprises the following steps: collecting network multi-dimensional data and generating a standardized network event set and a network control path input set; establishing a neural controlled differential equation model, and generating a continuous time context representation set through hidden state evolution; establishing a neuro-hox process identification model, performing intensity function modeling, and generating an event intensity prediction sequence set and an identification intermediate representation set; forming an intrusion behavior decision rule set and a reasoning configuration set through joint training; new network multi-dimensional data are collected, the reasoning configuration set operation model is loaded, and a new event intensity prediction sequence set and a new candidate trigger time set are output; and generating a network intrusion behavior recognition result set in combination with the intrusion behavior decision rule set. According to the method, time modeling and logical reasoning are fused, and high-precision intrusion identification is realized.

The invention discloses an intelligent port full life cycle management method and system based on digital twinning, electronic equipment and a storage medium, and belongs to the technical field of intelligent port digital and intelligent management, and the method comprises the steps: constructing a digital twinning model, and adjusting the digital twinning model to obtain an optimized digital twinning model; fusing and correcting to obtain a candidate state updating model; carrying out accuracy verification, and outputting an updated digital twinborn model when a verification error meets a preset convergence condition; identifying a maintenance demand and / or an operation optimization point, and generating a preliminary management decision suggestion through a decision rule engine; and performing simulation verification on the preliminary management decision suggestions, performing adjustment and optimization until a verification result meets a preset effect judgment condition, and outputting executable management decision parameters. According to the method, the full-life-cycle digital twin model is established for port planning, construction, operation and decommissioning stages, so that port cross-stage data and evaluation chains are kept continuous, and systematicness and consistency of full-life-cycle management are improved.

The invention provides a hospital financial auditing method and system based on multi-agent cooperation, and relates to the technical field of artificial intelligence, and the method comprises the steps: constructing a financial field knowledge base; obtaining an auditing demand of a user; in combination with a financial field knowledge base, converting the hospital financial related data from a natural language into an SQL language through an AI-SQL fusion engine; based on the auditing demand, through a workflow arrangement engine, auditing task distribution is carried out on the multiple agents; calling an AI-SQL fusion engine, and extracting input data required by each agent; based on the input data, executing an auditing task through each agent; based on the audit result of each agent, evaluating a risk level, and generating a risk evaluation result; and integrating the audit result and the risk assessment result, and determining an audit decision result according to a decision rule.

The invention relates to the technical field of retrieval enhancement, in particular to a hierarchical knowledge tree construction method of retrieval results and a program product. The invention provides a hierarchical knowledge tree construction method for retrieval results. The method comprises the following steps: obtaining a target problem and a retrieval result comprising a plurality of associated texts and corresponding text sources; clustering all vectors corresponding to all the associated texts to obtain multiple groups of candidate merged pairs; combining the plurality of associated texts through a large language model based on a preset mixed decision rule to obtain a combined text, and reserving a combined source; a hierarchical knowledge tree with a target problem as a root, each group of candidate text pairs as a subject layer, a combined text after combination and an independent associated text as sub-subject layers, a text source and a combined source as evidence layers and a conclusion corresponding to the target problem as a fact layer is constructed through a large language model; clear entities, relations and evidence paths are formed, and the reasoning ability and reasoning interpretability of the large language model are improved.