75results about How to "Improve results" patented technology

The invention relates to the technical field of data matching, in particular to an information matching management system and method for financial marketing platform data, and the system comprises a data acquisition module, a data analysis module, a dynamic adaptation module, a marketing strategy module, an anomaly detection module and a strategy implementation module. According to the method, behaviors and demands of different customer groups can be deeply understood through comprehensive analysis of the market data set and clustering analysis, and such statistical analysis can provide a scientific basis for making a marketing strategy, so that a financial institution can identify and utilize market trends and modes on the basis of data driving; according to the invention, the market change is predicted and quickly responded, and the improvement of the risk assessment result and the anomaly detection capability not only optimizes the resource allocation, but also enhances the control of potential risks, improves the overall efficiency and effect of marketing activities, enhances the accuracy of data-driven decisions, improves the customer service experience and brand loyalty, and promotes the sales growth.

The invention relates to the technical field of livestock and poultry breeding, in particular to an intelligent duck farm environment management optimization method and system, and the method comprises the steps: collecting real-time growth state information of breeding duck growth, employing a sensor to optimally collect environment influence factors concerning a breeding duck farm, and executing association to obtain the growth state information after information synchronization; filtering high-weight environment influence factors based on growth attributes in the growth state information blocks, and setting key environment influence factors of a plurality of growth scenes based on the high-weight environment influence factors; environment influence factor collection of a plurality of growth scenes is executed, a matching environment log of healthy growth of the breeding ducks is configured to execute semantic processing, and the similarity with the matching environment log is collected; and optimizing the breeding duck growth scene based on the similarity to obtain priority optimization entries, and optimizing the environment of the concerned duck farm. According to the method, the optimal value can be judged according to the duck breeding state data of the intelligent duck farm and the mixed breeding state of the breeding ducks, and the utilization efficiency and the output result of the duck farm are effectively improved.

The invention discloses a circuit logic reconstruction and process mapping method and device based on structure selection. The method comprises the following steps: reading and analyzing target design circuit information; one-to-one mapping and node collection; node priority cutting enumeration and maximum fan-out free cone calculation are carried out; calculating candidate equivalent nodes; creating a hybrid structure selection network; and performing logic reconstruction or process mapping. According to the method, related advantages and characteristics of structure selection are utilized, the structure advantages of various logic expressions are effectively combined, and various logic optimization strategies and process mapping are combined, so that the problem of structure deviation in a logic synthesis stage is solved, and the result quality of logic reconstruction and process mapping is improved. According to the method, logic optimization and process mapping are effectively combined, the performance is excellent, logic optimization and mapping results are obviously improved, a good experimental result is shown, a new research idea is provided for logic reconstruction and process mapping in logic synthesis, and the method has high engineering practical significance and application value.

The invention discloses a query method and device based on a natural language and data knowledge and a storage medium, and the method comprises the steps: constructing a first database based on collected metadata and corresponding data consanguinity, and constructing a second database based on collected historical business data; determining a plurality of data consanguinity corresponding to the metadata in the first database, and determining a plurality of joint knowledge vectors based on the metadata and the corresponding data consanguinity; the first query vector is matched with the multiple joint knowledge vectors, a target joint knowledge vector corresponding to the first query vector is determined, the first query vector corresponds to a first query text of the terminal equipment, and the first query text is natural language text information; generating a first SQL language instruction based on the target joint knowledge vector and a pre-trained generative language model; and querying in a second database by utilizing the first SQL language instruction, and obtaining and returning corresponding first business data to the terminal equipment.

The invention discloses an automatic causal structure generation method based on semantic representation and logical reasoning of a large language model. The method comprises the following steps: acquiring an input text; performing semantic coding and clustering on the obtained input text by utilizing a large language model, and establishing a candidate causal variable set; causal relationship detection is carried out on the established candidate causal variable set based on anti-fact intervention and do-calculation; performing causal direction judgment, and generating a directed acyclic causal graph meeting logic consistency; and on the basis of the generated directed acyclic causal graph, natural language interpretation is generated by using a large language model, and logic consistency closed-loop verification is carried out. According to the method, automatic generation from the natural language to the causal structure is realized, the causal variable set is automatically extracted and constructed from the unstructured natural language text, the defects that variables need to be manually defined and modeling depends on field experts in the existing causal modeling process are avoided, and the labor cost and professional threshold of causal structure construction are remarkably reduced.

The invention discloses a CVE-ATTamp (Cascade Vector Enhanced ATTamp) for a network security scene. The invention discloses a CK association fusion RAG construction method, and relates to the technical field of network security. The method comprises the following steps: (1) carrying out structured analysis and vectorization on CVE vulnerability information data, extracting features and fusing CVSS scores to carry out semantic embedding; (2) ATTamp; semantic modeling of a CK threat matrix is carried out, and a threat semantic space containing tactical level information is constructed; (3) heterogeneous knowledge association and graph construction: calculating cross-domain semantic similarity and establishing logic mapping between vulnerabilities and tactics in combination with expert rules to form a heterogeneous knowledge graph; and (4) attack logic-oriented RAG retrieval and generation: retrieving a key attack path sub-graph based on a hybrid scoring mechanism, and driving a large model to generate a security analysis report with logic depth. According to the method, attack logic is stored through mapping, so that the problems that the traditional RAG lacks inference ability and security data fragmentation are solved, and the accuracy and tactical value of automatic threat analysis are remarkably improved.

The invention discloses a monitoring method and system for the operation condition of transfer station leachate treatment equipment, and relates to the technical field of leachate treatment. The method comprises the following steps: acquiring equipment state information of each leachate treatment equipment, establishing an initial three-dimensional image model and a real-time three-dimensional image model of each leachate treatment equipment according to the equipment state information, and obtaining a real-time corrosion degree according to a difference degree between the initial three-dimensional image model and the real-time three-dimensional image model; calculating a specific data change rate according to specific data before and after the percolate passes through each percolate treatment device, and establishing a treatment effect-corrosion degree relation model; establishing a plurality of process nodes according to the real-time leachate treatment data, and further establishing a leachate treatment process and an equipment association network; and the equipment association network is associated with the treatment effect-corrosion degree relation model, so that the working state and the running state of the leachate treatment equipment are judged, and intelligent supervision on the running state of the leachate treatment equipment is realized.

The invention relates to the field of crop cultivation methods, and discloses a blueberry efficient substrate cultivation method suitable for southern districts, which comprises the following steps: preparing a series of nutrient solutions suitable for the growth and development of blueberries in southern districts, the series of nutrient solutions comprising a growth period nutrient solution, a flowering and fruiting period nutrient solution and a fruit swelling period nutrient solution, different nutrient solutions are provided for blueberry plants in different growth and development stages through a drip irrigation system, vegetative growth and reproductive growth balance of blueberries is promoted, and meanwhile high-quality and high-efficiency production of the blueberries is achieved in combination with matched cultivation and management technologies such as heavy retraction pruning and pest control and prevention and control. Compared with the prior art, the method has the advantages that balanced nutrients are provided for each growth and development stage of blueberry plants, the organic fertilizer solution is supplemented in the fruit expanding stage, the matched cultivation management technology is applied, blueberry crowns are transformed and updated, high-temperature harm is resisted, the number of branches of the plants is increased, the fruit part is increased, the fruit sugar content is increased, and the fruit yield and quality are obviously improved.

The invention belongs to the field of bidding and tendering, relates to a data analysis technology, is used for solving the problem that various requirements of enterprise users cannot be met due to the fact that bidding project recommendation is provided for the enterprise users according to a single standard or basis in the prior art, and particularly relates to an enterprise competition management method and system based on bidding and tendering big data. The enterprise management platform is in communication connection with a document screening module, a user analysis module, a recommendation output module and a bidding database; the document screening module is used for performing preliminary screening analysis on the bid invitation documents: marking enterprise users performing bid invitation analysis as analysis objects, and extracting preliminary screening items from the bid invitation database through the analysis objects; according to different target tendency and operation tendency combinations selected by the enterprise user, secondary screening is carried out on the primarily screened items, the matching degree of the recommended items obtained through secondary screening and the requirements of the enterprise user is high, and therefore the auditing efficiency of the enterprise user and the selection probability of the recommendation result are improved.

The present invention relates to a quality detection device for oral liquid of Chinese patent medicine, which integrates pretreatment, infrared spectrum, fingerprint spectrum, ultraviolet visible spectrum detection module and cloud analysis feedback system. Intelligent pretreatment is used to ensure sample suitability, multi-spectral technology is used to comprehensively analyze components, cloud database and random forest algorithm are combined, and types are quickly subdivided according to absorbance, peak position shift and area change are quantified in detail, and finally the quality grade is determined and feedback is given to users in real time. The system improves detection efficiency and accuracy, and strengthens the intelligent level of quality control of Chinese medicine.

This invention discloses a method, equipment, and medium for super high-rise construction monitoring based on digital twins, relating to the field of digital twin technology. The method includes: acquiring construction operation data from the super high-rise construction site and constructing a digital twin construction model; updating the boundary conditions of the digital twin construction model through a set of construction disturbance events to generate a disturbance-driven twin model; verifying the consistency between the virtual and real data of the model simulation state records in the construction state set and the disturbance-driven twin model to generate a twin credibility evaluation set; improving the local monitoring resolution of high-risk areas to generate twin resolution configuration results; performing local high-precision monitoring of the credible construction twin based on the twin resolution configuration results; and outputting super high-rise construction monitoring results. This invention, by improving the local monitoring resolution and generating twin resolution configuration results, avoids the waste of computational resources caused by global high-precision monitoring, thereby improving the efficiency of monitoring resource utilization and the response accuracy of high-risk areas.

This invention discloses a data management method and electronic device for a storage device, belonging to the field of storage technology. The data management method includes: when the lifetime of the hold table is less than the lifetime of the hold data area, writing data from the data cache to non-volatile memory, and the storage device entering a hibernation mode; when the lifetime of the hold table is greater than the lifetime of the hold data area, determining whether the data size in the data cache has reached a flush threshold; if the data size reaches the flush threshold, writing the data to non-volatile memory, and the storage device entering a hibernation mode; if the data size does not reach the flush threshold, saving the data to the hold data area, recording the boundary marker information of the data in the hold data area, and writing the system management data table to be migrated from the hold table to the non-volatile memory, and the storage device entering a hibernation mode. The data management method and electronic device for a storage device provided by this invention can extend the lifespan of the storage device.

This invention discloses a multi-channel electromagnetic induction detection method and system based on uniform magnetic field constraint. A uniform magnetic field is constructed within the detection area of ​​the structure under test. Electromagnetic induction response signals are acquired through multiple channels under the constraint of the uniform magnetic field, and the spatial coordinates of each acquisition channel are marked. The raw electromagnetic induction signals acquired by each channel are preprocessed, including noise reduction filtering, envelope extraction, and amplitude normalization. A deep learning model is then used to extract structural features from the preprocessed multi-channel electromagnetic response signals. Based on the spatial position information of each acquisition channel, magnetic field uniformity parameters, and response characteristics, a channel correlation weight matrix is ​​constructed to describe the spatial consistency and physical coupling relationships between channels. A multi-channel joint optimization model is constructed, and the equivalent electromagnetic parameter distribution inside the structure under test is reconstructed through iterative solution. This invention improves the reliability, repeatability, and engineering applicability of electromagnetic induction detection results.

The invention provides a spatial transcriptome tissue specific expression gene identification method and system based on a multi-dimensional statistical index, and belongs to the technical field of gene identification, the method comprises the following steps: obtaining gene expression data corresponding to a plurality of genes in different cells, different cells belonging to different types of tissues; the gene expression data is data processed based on a space transcriptome; for each gene, determining a basic statistical feature and a differential expression feature corresponding to the gene based on the gene expression data corresponding to the gene in different cells; determining specific index characteristics of the gene based on the basic statistical characteristics of the gene; and determining a tissue specific expression gene from the plurality of genes based on the basic statistical characteristics, the differential expression characteristics and the specific index characteristics corresponding to each gene. According to the invention, the accuracy of tissue specific expression gene recognition can be improved.

The application discloses a protease catalytic type attribution method based on protein cleavage site embedded vectors, comprising the following steps: constructing a cleavage site reference set of known protease substrates; obtaining N reference site vectors and M attribution site vectors; performing standardization processing on the N reference site vectors and the M attribution site vectors; calculating M*K first similarities and M*N second similarities; obtaining the nearest neighbor reference sites corresponding to each attribution site; obtaining the nearest neighbor support labels and the nearest neighbor support degrees of each attribution site; obtaining a credibility threshold required for attribution of the catalytic type of the attribution site; comparing the first similarity, the similarity difference and the nearest neighbor support degree with the credibility threshold required for attribution of the catalytic type of the attribution site; generating a corresponding credibility level or marking the attribution site as an uncertain candidate site; and the application can output attribution results of each attribution site according to the same data structure.

The embodiment of the invention provides a content playing progress processing method and a related device, and the method comprises the steps: responding to a first sliding operation for a playing progress bar of target content in a content playing page, entering a progress bar page from the content playing page, and displaying a preset progress bar corresponding to the playing progress bar and the first time point of the target content in the progress bar page, wherein the display precision of the preset progress bar is higher than that of the playing progress bar. And in response to a second sliding operation for the preset progress bar, displaying the progress bar change content of the preset progress bar and a second time point, different from the first time point, of the target content in the progress bar page. And when the sliding of the second sliding operation is finished, returning to the content playing page from the progress bar page, and displaying the playing content of the target content corresponding to the second time point in the content playing page. According to the method, the content playing progress can be more accurately skipped, so that the skipping result and the skipping experience of the content playing progress are improved.

The invention discloses a power failure customer appeal quantity prediction method based on data driving, and relates to the technical field of intelligent power grids. Comprising the following steps: comprehensively considering a plurality of inherent characteristics of a power failure line, objective dynamic factors in a power failure period and artificially determined factor components in a sample set construction process, classifying independent lines through an optimal clustering result based on line characteristics, and effectively dividing a sample set according to the classification result; carrying out expansion transformation on the divided sample set, so that the input characteristics of the training sample contain time sequence information, and the output result implies the time sequence change characteristics; and constructing a neural network model for each type of expansion sample set differentiation. The method can predict the customer power failure appeal changing along with time after power failure, and has higher reliability compared with the prior art.

The application provides an aluminum pad structure and a manufacturing method thereof. The manufacturing method of the aluminum pad structure comprises the following steps: sequentially forming a first aluminum layer, a protective layer and a passivation layer on a substrate; performing dry etching on the passivation layer by using a plasma containing fluoride ions to form a passivation layer opening, wherein the passivation layer opening is located above a preset region of the first aluminum layer and the bottom surface of the passivation layer opening stays on the surface of the protective layer; then forming a second aluminum layer on the protective layer and removing the part of the second aluminum layer which is higher than the passivation layer, thereby obtaining a sandwich-type aluminum pad structure composed of the first aluminum layer, the protective layer and the second aluminum layer. The manufacturing method of the aluminum pad structure can avoid the direct contact of the etching process with the aluminum pad, thereby eliminating the generation of aluminum fluoride compound residues and fluoride ion residues on the surface of the aluminum pad from the root. The aluminum pad structure is in a sandwich type, has less aluminum fluoride compound residues and fluoride ion residues, is beneficial to improve the bonding and yield test results, and reduces fluorine precipitation.

The invention provides a strabismus screening system based on multi-dimensional dynamic eye movement tracking. The strabismus screening system comprises a data acquisition module, a data preprocessing module, a data analysis module and a result feedback module. A cartoon experimental normal form is adopted for an experiment through a domestic Pupil.IO AIO eye tracker, so that the eye movement track is monitored, the stability of a fixation point is dynamically quantified, the dynamic change during jump reaction is captured, multi-dimensional eye movement feature calculation such as sight line difference analysis is carried out, and finally dichotomy result judgment is carried out. The problems of inherent limitation, high equipment price, insufficient clinical suitability of a data analysis model and lack of a standardized process of a traditional inspection method are effectively solved.

The embodiment of the application provides a kind of based on three-dimensional point cloud's corrosion depth distribution extraction method and related device, method includes: first, the three-dimensional point cloud data of the outer surface of the steel wire to be measured is obtained, and curved steel wire is converted to straight coordinate system, the radial distance and ring angle of each surface point are calculated;Then segmented point cloud is carried out according to preset interval along axial direction, in each segment, based on the median and median absolute deviation of outer deviation, threshold is set, the abnormal point located outside the cross section fitting circle is removed, and the real point in the corrosion pit is retained;Then, the uncorroded section at both ends is set as reference section, the radius sequence is fitted, and the median value is taken as global reference radius;Finally, for the retained real point, the absolute corrosion depth is calculated, and the corrosion depth distribution probability density diagram and two-dimensional heat map are output, the interference problem of scanning noise, registration error and outside fly point to corrosion depth statistics can be solved, and the data purity and result reliability of corrosion depth calculation are significantly improved.

The application discloses a kind of distribution outside data delay calibration method and system, the method is first obtained historical layout training set and is based on line length threshold value identification distribution outside sample subset, then utilize inverse probability weighting method to calculate the calibration weight of each sample in training set, wherein distribution outside sample obtains higher weight, then using the weight to training set is weighted and re-trains machine learning model, obtains the delay prediction model after calibration, after calibration model deployment, directly used in interconnection delay prediction in engineering change order layout;The application is intelligently weighted and weighted training to internal sample of historical training set, without relying on any new tape data, effectively corrects the systematic underestimate of model to distribution outside long interconnection delay, so as to greatly improve the prediction accuracy, avoid the buffer insertion omission and timing violation caused by prediction deviation, enhance the efficiency and reliability of timing convergence.

The present invention relates to the field of data matching technology, specifically to an information matching management system and method for financial marketing platform data, the system comprising a data acquisition module, a data analysis module, a dynamic adaptation module, a marketing strategy module, an anomaly detection module, and a strategy implementation module. In the present invention, by comprehensively analyzing market data sets, cluster analysis helps to more deeply understand the behaviors and needs of different customer groups. Such statistical analysis can provide a scientific basis for the formulation of marketing strategies, enabling financial institutions to identify and utilize market trends and patterns on a data-driven basis, predict market changes and respond quickly, and improve risk assessment results and anomaly detection capabilities, not only optimizing resource allocation, but also enhancing control over potential risks, improving the overall efficiency and effectiveness of marketing activities, enhancing the accuracy of data-driven decision-making, improving customer service experience and brand loyalty, and promoting sales growth.

The application relates to the cross technical field of information security and artificial intelligence, and provides a user abnormal behavior detection and threat identification method based on multi-entity association. In order to solve the problem of how to realize robust detection and threat identification of user abnormal behavior in a multi-entity environment without relying on normal behavior boundary modeling, the multi-entity operation relationship in the security log is aggregated into a heterogeneous behavior graph according to a time window, is encoded into a behavior memory state vector through a relationship graph attention network, is modeled into a long-term behavior memory track through a time sequence convolution network, and then active disturbance is applied and the recovery dynamics characteristics of the reconstruction error after disturbance are observed. The extracted recovery dynamics characteristic parameters are input into a classification network for abnormality determination, finally, the abnormality signal is propagated along the entity association edge, and the threat identification result is output based on the semantic matching of a threat tactic knowledge graph, so that the detection accuracy and robustness when facing advanced camouflage attacks are significantly improved.

This invention discloses a 3D drawing similarity retrieval method based on a spatial assembly structure vector tree, belonging to the field of 3D information retrieval technology. The method includes: constructing a spatial assembly structure vector tree based on a set of node spatial attributes and a set of assembly constraint attributes, according to the parent-child relationship of the original assembly hierarchy tree; performing normalization and standardization processing to generate a constraint-enhanced assembly vector tree; extracting hierarchical features from the constraint-enhanced assembly vector tree to generate hierarchical retrieval fingerprints, and matching these fingerprints with the hierarchical retrieval fingerprints of historical 3D drawings to establish hierarchical candidate mapping relationships and generate a hierarchical candidate retrieval spectrum; based on the hierarchical candidate retrieval spectrum, performing hierarchical flexible alignment to generate a flexible alignment mapping set, and calculating the comprehensive similarity between the 3D drawing to be retrieved and each historical 3D drawing. This invention achieves continuous organization of hierarchical correspondences and fine discrimination of similar features, improving the accuracy and usability of the retrieval results.

The present invention discloses a method and system for intelligent detection of water conservancy gate state, the method includes data collection, data preprocessing, gate opening and closing detection, water level and water pressure detection, mechanical wear detection and water conservancy gate state detection. The present invention relates to the field of gate state detection technology, specifically to a method and system for intelligent detection of water conservancy gate state, the scheme adopts the analysis of three states of gate opening and closing state, water level and water pressure state and mechanical wear state, analyzes short-term decision and early warning through gate opening and closing state and water level and water pressure state, and analyzes long-term maintenance and use through mechanical wear state, so as to perform multi-faceted and multi-angle intelligent gate state detection; adopts a bidirectional temporal convolutional network method combined with convolutional feature extraction to perform water level and water pressure state detection; adopts an integrated model combining gated cyclic unit and recursive neural network to perform mechanical wear state detection.

The present application relates to the technical field of equipment fault positioning, and more particularly to a power system equipment fault positioning method based on chain penetration technology, and the present application proposes the following scheme: obtaining a target alarm event and process variable data in a corresponding time window; constructing an industrial knowledge graph based on equipment topology relationship, process flow direction relationship, control interlocking relationship and historical fault knowledge; performing multi-level correlation traversal, correlation screening and priority evaluation around the target alarm event to determine a candidate measurement point node set; combining real-time measurement value and prediction value deviation, correlation distance, historical co-occurrence frequency and propagation direction consistency to calculate reason contribution scores, and outputting a potential fault reason set and troubleshooting suggestions. The present application improves the positioning accuracy and interpretability of abnormal alarms of complex industrial equipment.

The present invention discloses a method and system for analyzing civil tracks based on hybrid vector retrieval enhancement. The method includes: obtaining the location information and description information of multiple known targets and constructing a dual-vector feature library; obtaining the location information and description information of the target to be analyzed and constructing a dual-vector query feature; the dual-vector query feature includes a geographic vector query feature and a semantic vector query feature; similarity calculation is performed based on the dual-vector query feature and the dual-vector feature library to obtain the final similarity between the target to be analyzed and the known target; enabling the large model service, initializing the large model using the system prompt word, and obtaining the large model question-answering framework; inputting the final similarity calculation result into the large model question-answering framework to obtain the analysis result output by the large model question-answering framework. The present invention is applicable to the fields of maritime navigation, logistics and transportation, and can perform interactive and friendly accurate analysis and interpretation of civil track data.

The application discloses an optimization method of a rolling wing blade connecting rod mechanism, and belongs to the field of rolling wing aircrafts.The blade swing angle is represented by structural parameters of the blade connecting rod mechanism, and the structural parameters of the blade connecting rod mechanism, including the lengths of the connecting rods, the coordinates of key hinge points and configuration bias, are defined as a variable vector; multiple target functions are constructed for multiple key working conditions such as hovering, forward flight and transition, the multiple target functions not only quantify the performance of the connecting rod mechanism itself, such as kinematic fitting error with a target motion law, dynamic stability index and force transmission efficiency index, but also cover aerodynamic efficiency index and aerodynamic acoustic index which are directly related to the performance of the whole machine; and the optimal structural parameters of the blade connecting rod mechanism are obtained by solving the multiple target functions under constraints, so that the global optimization of the structural parameters of the blade connecting rod mechanism under multiple working conditions, multiple targets and multiple constraints can be realized.

The invention discloses a semi-physical simulation system of automatic cycle control, which relates to the technical field of semi-physical simulation in ballistic missile simulation technology, including automatic cycle test software, Higale real-time simulation system, inertial measurement combination simulator, integrated navigation simulator, flight control computer, steering gear system and program-controlled power supply system; the automatic cycle test software and the Higale real-time simulation system are connected via TCP / IP protocol to automatically complete test parameter binding and test result acquisition, and the Higale real-time simulation system is deployed with a six-degree-of-freedom dynamics module, an inertial measurement combination simulator and an integrated navigation simulator. The invention reduces the impact on time and accuracy brought by human operation, greatly improves test efficiency, thereby improving the coverage of test parameter states and the credibility of simulation results, and reducing the hidden dangers of flight tests.

This invention relates to the field of computer data processing technology, and in particular to an optimized configuration method and electronic device for a wind-solar-hydrogen production system. The method includes constructing an upper-level model and a lower-level model; using the minimum unit number of each subsystem of the wind-solar-hydrogen production system as the decision variable of the upper-level model; acquiring basic data, generating an initial population of decision variables based on the upper-level model, and transmitting it to the lower-level model; clustering the wind and solar power output based on the lower-level model, reconstructing it into typical annual data, solving for the annual data of the wind-solar-hydrogen production system, and outputting it to the upper-level model; calculating the annual hydrogen production cost based on the upper-level model; determining whether a preset genetic algorithm termination condition is met; if so, outputting the result; otherwise, performing crossover and mutation on the decision variable population and transmitting it to the lower-level model. Therefore, this invention can solve the problems of existing technologies being unable to guarantee the reliability requirements of continuous hydrogen supply from coal chemical plants and being unable to assess the capacity gap of hydrogen production systems under extreme and continuous severe weather conditions.