292 results about "Early identifications" patented technology

The invention belongs to the technical field of coal mine risk early warning, and particularly relates to a coal mine goaf multi-risk comprehensive early warning method and system based on machine learning, and the method comprises the steps: collecting mine pressure, gas and hydrological real-time data in real time through a multi-temporal-spatial-scale sensor, and obtaining a dynamic coupling relation basic data set based on the real-time data; preprocessing noise and missing values according to the dynamic coupling relationship basic data set, and modeling node connection between a geological structure and mine pressure change by adopting a graph neural network to obtain space-time heterogeneous feature representation; non-linear features are analyzed through spatial-temporal heterogeneous feature representation, and a multi-scale dynamic mode is determined; acquiring a risk conduction path in the multi-scale dynamic mode, and acquiring an early recognition signal of a potential disaster chain; based on the early recognition signal, a long-short-term memory network is used for processing a sequential sequence, and the probability of the compound disaster is judged; a high-risk area is extracted from the composite disaster probability, and real-time early warning model parameters are obtained; and generating alarm output according to the real-time early warning model parameters.

The invention provides a mine abnormal event real-time identification method and system based on time sequence characteristics, and relates to the technical field of mode identification, and the method comprises the steps: carrying out the time-space alignment and semantic annotation of multi-modal monitoring data, and constructing a time sequence knowledge graph; calculating a dynamic association weight between entities, and analyzing a risk propagation path; predicting a risk situation based on a historical evolution rule; and dynamically generating a differential early warning strategy and establishing a closed-loop tracking system. According to the invention, early identification, accurate prediction and efficient disposal of mine safety risks can be realized, and the mine safety management level is improved.

The invention discloses a crop disease diffusion prediction method and system based on multi-modal fusion, and the method comprises the following steps: S1, collecting and preprocessing an RGB image sequence and a sensor data sequence of a crop growth environment, and generating an RGB image time sequence difference result and a sensor difference result through time difference processing; s2, mapping the RGB image time sequence difference result and the sensor difference result to a shared time sequence space through a time alignment algorithm, and generating a sensor alignment result and an RGB alignment result; s3, an FD-ViT prediction model is constructed; inputting the sensor alignment result and the RGB alignment result into an FD-ViT prediction model for prediction, and generating a prediction result; and S4, generating a disease diffusion thermodynamic diagram and early warning information according to a prediction result. According to the method, RGB image data and sensor network data are fused, a Transform-based time sequence prediction model is constructed, and early recognition and diffusion trend prediction of crop diseases are realized.

The invention discloses a multi-agent large-model disease diagnosis knowledge reasoning system based on data dual drive, and relates to the technical field of artificial intelligence assisted medical diagnosis. The system collects patient symptom follow-up records, laboratory test results, observation diagnosis probabilities and expert diagnosis recommendation results in a multi-source manner; time sequence evolution characteristics are extracted, a time sequence diagnosis sensitivity coefficient is calculated, and early recognition of disease risks is achieved; in combination with anti-fact simulation and statistical reasoning, a causal consistency coefficient is obtained and is used for verifying causal reasonability of observation diagnosis and contrast results; based on agent group consensus analysis, calculating a game consistency coefficient for judging the credibility of a diagnosis conclusion; positioning and multi-level verification are carried out on abnormal reasoning steps and knowledge fragments, so that the reliability and safety of a result are guaranteed; continuous optimization of the diagnosis model is realized through a log analysis and knowledge backflow mechanism; according to the invention, the accuracy, interpretability and safety of disease diagnosis can be obviously improved.

The invention discloses a coronary artery calcification early warning system for type 2 diabetes patients, and relates to the technical field of medical detection. A data acquisition module is used for acquiring continuous physiological parameter data of a user; the risk modeling module is combined with coronary artery calcification evolution characteristics in historical clinical samples to construct a multi-parameter dynamic association model; an index weight calculation unit generates a risk influence factor vector based on a sensitivity analysis result of the physiological indexes on risk prediction; the machine learning analysis module performs iterative training on the prediction model by adopting an integrated learning algorithm, and performs prediction updating by utilizing a risk influence factor vector; the early warning trigger module dynamically generates a graded early warning signal according to the grading trend and a set threshold value; the weak item positioning module carries out contribution degree analysis and anomaly recognition on the key risk indexes and automatically generates personalized intervention suggestions; according to the invention, early recognition and dynamic early warning of coronary artery calcification progress can be realized, and the method is suitable for intelligent early warning management scenes of chronic disease cardiovascular risks.

The invention belongs to the field of medical artificial intelligence, and particularly relates to a multi-modal data acquisition and fusion method for Alzheimer's disease. The method comprises the following steps: firstly, synchronously acquiring eye movement, expression, voice, gait and grip strength data of a subject through a virtual reality multi-task normal form, and combining with an MoCA scale to score a result; then preprocessing and feature extraction are carried out on each modal data, and unified feature representation is constructed; on the basis, a cross-modal attention mechanism is adopted to realize interaction and weighted fusion of multi-modal features, and a unified fusion feature vector table is generated; and finally, outputting structured data organized according to task fragments for auxiliary evaluation and modeling of cognitive impairment. The method can effectively solve the problems that in the prior art, single-mode information is insufficient, and multi-mode data are difficult to align and fuse, has the advantages of being low in cost, easy to popularize and high in detection accuracy, and can be widely applied to early recognition and auxiliary diagnosis of the Alzheimer's disease.

The invention discloses a switch cabinet latent fault remote diagnosis system, method, device and medium, and belongs to the technical field of fault diagnosis, and the system comprises a multi-dimensional sensor sensing module, a data transmission module, a data fusion analysis module, and a data storage and fault detection module. The multi-dimensional sensor sensing module is used for acquiring multi-dimensional data of the operation state of the switch cabinet; the data transmission module is used for transmitting the multi-dimensional data to the data fusion analysis module; the data fusion analysis module is used for performing data preprocessing on the multi-dimensional data to obtain preprocessed multi-dimensional data, and determining a fault classification result according to the preprocessed multi-dimensional data; and the data storage and fault detection module is used for performing fault identification based on the real-time monitoring data and the fault classification result. According to the invention, early recognition and accurate diagnosis of the latent fault of the switch cabinet are realized, and the downtime and maintenance cost of equipment are reduced.

The invention relates to the technical field of remote sensing, and discloses a forest land health state analysis method and system based on multi-source remote sensing image analysis. The system comprises a multi-source remote sensing acquisition module, a feature extraction and fusion module, a health assessment module, a traceability analysis module, a strategy matching module, a visual reconstruction module, an early warning decision module and an execution feedback module, and constructs a multi-modal forest land observation data set by fusing multi-source remote sensing data of multispectrum, hyperspectrum, radar and thermal infrared. The comprehensive extraction of multi-dimensional information such as vegetation coverage, canopy biochemical characteristics, under-forest structures and surface thermal environments is realized, the limitation of single data source analysis is overcome, and the comprehensiveness and accuracy of forest land health condition evaluation are improved; through dynamic comparison of multi-stage remote sensing images and health risk level mapping, early identification and early warning of forest growth abnormity, degeneration trend and pest and disease risk are realized.

The invention relates to the technical field of synthetic aperture radars, discloses a method for monitoring activity of a high-order loose body based on multi-source SAR data, and aims to solve the problem that an existing high-order loose body activity identification mode is poor in accuracy. Obtaining multi-source SAR data and carrying out adaptive preprocessing; extracting a maximum back scattering feature and quantifying activity; carrying out activity identification and multi-scale feature analysis on the high-position loose body; and comprehensively evaluating the multi-dimensional risk of the movable high-position loose body. Through organic combination of multi-track collaboration, fine feature extraction and multi-dimensional model evaluation, an effective scheme is provided for solving the problems that early recognition of the high-position loose body in the alpine and valley region is difficult and risk evaluation is inaccurate, and the accuracy and timeliness of monitoring and early warning are remarkably improved.

The invention discloses a VTE real-time monitoring and intelligent prevention and treatment system, and belongs to the technical field of intelligent prevention and treatment, and the system comprises a baseline construction module which is used for collecting multi-dimensional VTE parameters, calculating the normal fluctuation interval of each parameter to form an initial individualized baseline, and constructing a self-adaptive individualized baseline through threshold calibration; the trend identification module is used for dynamically setting a sliding window duration, calculating a VTE trend slope and a VTE product deviation, constructing a trend constraint and a product deviation constraint, and when the two constraints are not met at the same time and the continuous deviation duration is exceeded, judging that the deviation is continuous abnormal deviation; the time sequence risk prediction module is used for calculating deviation values of various parameters of the target patient, constructing a space-time fusion feature matrix, inputting a time sequence risk prediction model and outputting a VTE risk probability; and the early warning and intervention module is used for performing double judgment and intervention, setting three-level early warning and intervention measures, calculating an improvement rate to verify a prevention and control effect in real time, and realizing VTE early recognition and intelligent prevention and control.

The invention discloses a black pig breeding disease intelligent monitoring management method based on big data, and relates to the technical field of animal husbandry intelligent management and animal disease monitoring, and the method comprises the following steps: obtaining multi-dimensional data information of pigs in a black pig breeding scene in real time through a sensor network, a video monitoring system and a physiological information collection device; the multi-dimensional data collected in real time is preprocessed and standardized, and the original data quality and the data analysis effectiveness are improved. According to the invention, by introducing an adaptive adjustment mechanism, the problem of pathological data loss caused by excessive elimination of abnormal values in the prior art is solved, and accurate retention and dynamic tracking of early disease signals of pigs are realized. The method integrates multi-dimensional feature extraction and intelligent evaluation, has pathological trend perception and processing strategy adaptive adjustment capabilities, effectively improves the early recognition sensitivity and discrimination accuracy of a disease early warning system, and provides more scientific health management support for farms.

The invention discloses a building defect recognition method and system based on image recognition, and particularly relates to the technical field of computer vision. Comprising a data acquisition and preprocessing module, a building defect knowledge graph construction module, a building defect identification module, a building defect quantitative analysis module, a building defect risk assessment module and a building defect risk early warning module. According to the method, the extracted building defect information and structured domain knowledge are deeply fused, the reason for forming the defect can be further reasoned, deep understanding and diagnosis of the defect are achieved, the interpretability of model output is remarkably improved, quantitative parameters, diagnosis causes, component importance and historical change trends of the defect are calculated, and the accuracy of the model output is improved. According to the method, the quantitative building defect risk index is obtained, early warning of different levels is automatically triggered according to the preset value, early recognition and predictive maintenance of the structure risk are achieved, and active early warning and risk prediction of the building defect are achieved.

The invention discloses a medical image automatic identification system based on a neural network, and relates to the technical field of medical image identification. The method is used for solving the problem that early recognition of neurodegenerative diseases is difficult due to medical image and genome data splitting and poor model interpretability in the prior art. The method comprises the following steps: firstly, extracting multi-scale features of a brain structure through a three-dimensional convolutional neural network and a self-attention mechanism, calculating a multi-gene risk score based on a risk site, and encoding the score into a feature vector; secondly, using a cross attention mechanism to take gene features as query vectors, fusing the gene features with image features, and generating brain structure anomaly features under gene regulation; then, gradient weighting class activation mapping is applied to generate a visual thermodynamic diagram, and gene-image association weight weighting is combined to construct a brain region risk distribution diagram; and finally, a high-risk brain region space coordinate set is extracted through threshold segmentation, and an accurate quantification basis is provided for early recognition.

The invention discloses an elevator operation and maintenance full-life-cycle fault monitoring and diagnosing method and system, and relates to the technical field of data analysis, and the method comprises the following steps: collecting fault action data of known elevator component parts, and establishing an elevator component part fault tree map; marking a historical operation process instruction of the target elevator equipment, analyzing the matching between the historical operation process instruction and the assembled building, and evaluating the current state of each component of the target elevator equipment; running data of all parts of the real-time target elevator equipment are obtained, similar fault screening is conducted on the running data and the known elevator component fault tree atlas, an undetermined fault set of all the parts of the real-time target elevator equipment is obtained, an elevator part-fault confidence coefficient function is established, and fault types of all the parts of the target elevator equipment are generated. The method has the advantages that the fault early recognition capability and the fault diagnosis accuracy are improved, the fault processing time is shortened, and the reliability of elevator equipment is improved.

The invention relates to the technical field of material detection, and discloses an identification analysis and method for recessive degradation of a flexible conductive composite material, and the system comprises a perturbation excitation module, a response collection module, a time sequence feature extraction module, a micro-response trend consistency analysis module, and a state evaluation and early warning module. Multiple periods of small-signal capacitance excitation are applied to the material, pi-pi stacking slippage and interface polarization behaviors in a conductive network are excited, electrical response data are collected to construct a time sequence feature vector, and the deviation degree of the evolution trend is further calculated. And based on the trend deviation index, performing health state scoring and triggering early warning, thereby realizing early recognition and evaluation of material recessive degradation.

The invention discloses a deep venous thrombosis early warning method and system based on multi-modal physiological signals, and relates to the technical field of health monitoring. The method is used for realizing thrombus early recognition and graded early warning. The method comprises the following steps: synchronously acquiring physiological signals of lower limbs on two sides through a multi-modal sensing device, performing motion artifact elimination and signal quality screening, and generating preprocessed time sequence data; performing dynamic time warping and feature extraction on the time series data, calculating a bilateral symmetry index and an instantaneous and accumulated time-varying difference degree, and fusing the bilateral symmetry index and the instantaneous and accumulated time-varying difference degree with absolute physiological parameters to form a comprehensive risk assessment vector; establishing a dynamic baseline based on the historical risk vector, calculating a current deviation index, and recognizing an abnormal state in combination with a double-side cooperation mode; and according to the deviation index and the duration, in combination with the abnormality type and the severity, generating a graded early warning signal containing a risk grade and a disposal suggestion through multi-stage judgment, thereby realizing early warning and risk management of thrombus.

The invention discloses a turning gear monitoring system and a control method thereof, and aims to solve the problem that an existing turning gear lacks comprehensive real-time monitoring and fault early warning. The system comprises a monitoring module, a data processing module, a control module and an early warning execution module, wherein the monitoring module, the data processing module, the control module and the early warning execution module are connected in sequence; the monitoring module is used for monitoring operation data of the turning gear; the data processing module is used for monitoring and diagnosing the turning gear by utilizing a preset health database according to the operation data and generating a diagnosis result; the control module is used for generating a corresponding early warning instruction based on preset graded early warning according to the diagnosis result; and the early warning execution model is used for executing a corresponding alarm action according to the early warning instruction. Through cooperative work of multiple modules, the system can integrate various operation parameters and mechanical state data, early recognition of complex faults is achieved, and the early warning efficiency of early faults of the turning gear is improved.

The invention relates to a feeding intolerance burst risk early warning method and device based on multi-modal dynamic monitoring. The method comprises the steps that a gastrointestinal tract peristaltic wave image sequence under the abdominal skin surface of a patient is collected; collecting borborygmus data; collecting physical sign physiological data of the patient; monitoring behavior and posture data of the patient in real time; tracking the infinitesimal displacement and deformation of the gastrointestinal wall in the gastrointestinal tract peristaltic wave image sequence, and extracting and quantifying the speed, direction and wave crest and trough change rate of the peristaltic wave and the regularity index of the peristaltic wave; calculating a physiological deviation degree between the fusion feature at the current moment and an expected normal mode generated by the physiological feature baseline model of the patient; and inputting the physiological deviation degree, the change rate of the physiological deviation degree in the time dimension and the accumulated deviation duration into a conditional variation risk assessment model, and outputting a potential spatial distance. Through the non-invasive and highly quantified feeding intolerance risk model, the early recognition capability and clinical intervention efficiency of feeding intolerance are remarkably improved.

The invention discloses a rapid progression type nasopharyngeal carcinoma risk prediction method based on an artificial neural network, and relates to the field of medical informatics crossing. The invention provides a rapid progressive nasopharyngeal carcinoma risk prediction method based on an artificial neural network, and aims to solve the problem that a rapid progressive nasopharyngeal carcinoma patient is difficult to recognize in time by depending on TNM staging and experience judgment in the prior art. According to the method, historical case data collection, missing value filling and standardization preprocessing, core feature determination through feature screening, class imbalance correction, feature coding and feature matrix construction are sequentially carried out, an artificial neural network model is trained and optimized under a cross validation framework, and performance and threshold values are determined on a validation set. During clinical application, patient features are input, and the model outputs a rapid progress risk probability and a risk level. Compared with a conventional staging or linear model, the method can improve the prediction accuracy, and achieves the early recognition and individualized treatment of a high-risk patient.

The invention relates to the technical field of biomedical engineering, and discloses a VTE system execution method and system combined with a standardized scale, and the method comprises the steps: carrying out the standardization processing of original data, and obtaining the standardized data; constructing a risk intensity flow; carrying out integral operation to obtain an accumulated integral value of the risk factor; calculating a dynamic risk trajectory; fusing the baseline score value of the standardized scale with the dynamic risk trajectory to obtain a total risk trajectory; calculating an instantaneous risk probability, and judging whether VTE risk early warning is triggered or not according to a self-adaptive threshold value; according to the method, continuous quantitative monitoring of VTE risk factors is achieved by constructing the body position, anesthesia, hemostatic and circulation parameter four-dimensional risk intensity flow, early recognition and timely early warning of VTE risks are achieved through double mechanisms of continuous triggering and acute triggering, and the accuracy, the real-time performance and the individualized level of risk assessment are remarkably improved.

The invention provides an emergency call system, method and device based on AI vision and a storage medium, and relates to the technical field of active safety of intelligent network connection automobiles. The system comprises a camera module, a DMS module, an OMS module, an AI vision algorithm module, an AI vision application module and an eCall application module which work cooperatively. Aiming at the three defects that a traditional emergency call system depends on collision triggering, data dimensions are deficient and active protection is absent, DMS and OMS are innovatively and deeply integrated with the emergency call system, a multi-modal fusion decision mechanism is realized in combination with a layered AI visual architecture, and through deep cooperation of camera shooting, DMS, OMS, AI visual algorithm, AI visual application and an eCall application module, the multi-modal fusion decision mechanism of the emergency call system is realized. A full-link intelligent safety closed loop of sensing, decision making, control and rescue is constructed, early recognition, active intervention and accurate rescue of sudden diseases of drivers, serious fatigue disability and emergency physiological events of passengers are achieved, and meanwhile the rescue quality of collision accidents is improved.

The invention provides a construction electricity consumption and special equipment safety monitoring method and system, and relates to the field of construction safety management, and the method comprises the steps: carrying out the time sequence analysis, spatial analysis and logical reasoning of multi-source heterogeneous data containing electric power parameters and special equipment operation state parameters; processing the collected multi-source data based on a space-time sequence network to obtain a current health state score and a residual service life prediction value of the equipment; according to the analysis result of the time sequence dimension, the analysis result of the space dimension and the analysis result of the reasoning dimension, generating the severity of the abnormal condition of the monitored data, the occurrence reason of the abnormal condition and the affected upstream and downstream equipment as early warning information; and generating predictive maintenance early warning information according to the health state score and the residual service life prediction value. The early recognition and predictive maintenance of the performance degradation trend of the equipment are realized, and the intelligent level of construction electricity consumption and special equipment safety monitoring is remarkably improved.

The invention relates to the technical field of risk analysis, in particular to a water conservancy project risk assessment method based on big data, which comprises the following steps: acquiring deformation and water level signals and converting the signals into a sequence, generating an original monitoring sequence based on a timestamp, identifying a whole process span and intercepting and generating a periodic data set, generating a complete monitoring sequence by using interpolation filling, constructing a water level displacement closed curve, calculating a vector product difference value to obtain a geometric area, calculating an arithmetic mean value to generate a centroid coordinate, obtaining a dissipated energy increase coefficient and a structure permanent deformation cumulant, and generating an evaluation result based on threshold comparison. According to the method, the two-dimensional phase space model is established, the structure response is converted into the closed trajectory curve, the hysteresis loop area and centroid offset are calculated, energy dissipation and deformation accumulation of the structure under cyclic loading are quantitatively represented, abnormal plastic damage is effectively stripped, and early recognition of the potential irreversible damage risk of the structure is achieved.

The invention relates to an underground gas cut monitoring and early warning method based on multi-parameter fusion. The method comprises the steps that key parameters in well site data are collected, and standardized preprocessing is conducted on the key parameters to obtain standardized data; classifying and marking the standardized data as initial feature data, development feature data and danger feature data based on a gas cut evolution rule; constructing an underground gas cut monitoring and early warning model, and training the underground gas cut monitoring and early warning model based on the standardized data and the marks; inputting to-be-measured well field data into the trained underground gas cut monitoring and early warning model to obtain a judgment result which is output by the underground gas cut monitoring and early warning model and is classified as an underground gas cut grade; the problems of gas cut early recognition lagging and inaccurate grading in the drilling process are solved.

The invention discloses an early landslide disaster recognition method based on a Beidou unmanned aerial vehicle and remote sensing, and belongs to the technical field of geological disaster monitoring. The method comprises the following steps: firstly, cooperatively acquiring and preprocessing SAR images, Beidou GNSS and unmanned aerial vehicle multi-source data; performing time sequence coupling fusion; then, SAR deformation intensity, Beidou displacement dynamic trend, unmanned aerial vehicle surface fracture and three-dimensional terrain multi-source features are extracted by using a CNN network and fused; inputting the fusion features into a pre-constructed ConvNeXt-ViT double-path landslide early warning model, and outputting a landslide risk grade and a deformation rate in the next 15 days; and finally, closed-loop early warning within 10 minutes is executed through an integrated software and hardware early warning system. According to the landslide early warning method, macroscopic-mesoscopic-microscopic all-directional monitoring of the landslide is achieved, the cross-regional adaptability of the model is high, the early warning precision and timeliness are remarkably improved, the false alarm rate is smaller than 10%, and the missing report rate is smaller than 5%.

The invention discloses a lung function evaluation method for a patient in a pneumonia rehabilitation period, and relates to the technical field of rehabilitation management. Extracting a lung function feature set based on the image data; analyzing based on multi-dimensional parameter association, and establishing a dynamic mapping relationship between the lung function feature set and associated physiological parameters; identifying abnormal associated parameters through a dynamic mapping relation between the lung function feature set and the associated physiological parameters, and when the deviation between the image feature recovery degree and the corresponding physiological parameter recovery degree exceeds a preset threshold value, judging that functional abnormal parameters exist; and matching a potential complication database based on the abnormal parameter type, analyzing the relevance between the abnormal parameter and a preset disease feature model, and outputting secondary disease risk early warning. The method has the advantages that through cross-modal fusion of the lung image features and the dynamic physiological parameters, the early recognition precision of the pneumonia convalescence complication is remarkably improved.

The invention relates to the technical field of medical image processing and intelligent grading recognition of tumors, and discloses a head and neck squamous cell carcinoma early recognition method based on radiomics and deep learning, which comprises the following steps: acquiring head and neck CT (Computed Tomography), MRI (Magnetic Resonance Imaging) and PET (Positron Emission Tomography) three-mode image data; a two-stage dynamic structure alignment mechanism is adopted for registration; extracting fusion radiomics features; constructing a tumor sub-feature map; outputting a tumor level and a prediction confidence coefficient through the uncertainty prediction model; and generating a saliency interpretation heat map. In the prior art, a single-mode image or a superficial layer texture feature extraction model is depended on, and especially under the condition that a heterogeneity tumor region boundary is fuzzy and different modes have significant structure offset, high-confidence accurate discrimination of a real staging state of a tumor cannot be realized. According to the method, the accuracy of early stage identification of the head and neck squamous cell carcinoma is improved by introducing the significance guide registration mechanism and the improved graph convolutional neural network and combining Dropout reasoning.

The invention discloses a fire alarm inspection system and method based on big data, and relates to the technical field of intelligent fire protection and public safety. The problems that a traditional fire fighting system is serious in data island phenomenon, single in risk perception dimension, static and rigid in early warning model, high in false alarm rate and low in inspection efficiency are solved. According to the method, a fire-fighting theme database is constructed, and based on a multi-dimensional dynamic analysis model, comprehensive analysis of space aggregation risks, time sequence evolution trends and equipment linkage logic is realized, and dynamic fire danger hidden danger indexes are calculated; a key monitoring object is positioned through a visual thermodynamic diagram, grading alarm is triggered, differentiated inspection tasks and an optimal route are generated based on the intelligence, and emergency disposal linkage is supported; closed-loop management from global risk perception, intelligent analysis and early warning to precise inspection treatment is realized, and early recognition and early warning accuracy and resource scheduling efficiency of fire hazards are remarkably improved.

The invention provides a schizophrenia early recognition and risk prediction method and system based on multi-mode speech analysis, and belongs to the technical field of medical artificial intelligence. The method comprises the following steps: prospectively acquiring natural voice data of an individual in a clinical high-risk period, performing clinical follow-up visit for two years, and constructing a first Chinese schizophrenia clinical high-risk voice longitudinal data set; based on the data set, audio and text features in the voice are extracted, a layered multi-mode integrated model is constructed, semantic coherence and acoustic features are fused, and early prediction of the schizophrenia transformation risk is achieved; through feature importance analysis, key voice marks closely related to negative symptoms, thinking disorder and cognitive impairment are recognized, and non-invasive, accurate and interpretable early recognition is achieved. According to the method, the technical blank of early screening tools for schizophrenia in Chinese context is filled, and the accuracy and stability of early recognition are remarkably improved.

The invention discloses a forest fire early recognition and early warning system and method based on multi-source sensing and artificial intelligence, and belongs to the technical field of forest fire intelligent monitoring. The system comprises a gridding multi-source sensing deployment module, a multi-modal time sequence data collaborative acquisition module, a cross-modal feature coupling and state diagnosis module, a multi-source heterogeneous feature adaptive fusion and fire behavior identification module and an intelligent grading early warning and dynamic resource collaborative scheduling module. The method comprises the steps of heterogeneous grid division and sensing deployment, multi-modal time series data collaborative acquisition, cross-modal feature coupling analysis, multi-channel attention fusion recognition, and graded early warning and dynamic scheduling. Through multi-source data collaborative sensing, cross-modal feature coupling analysis, adaptive fusion recognition and intelligent scheduling response, accurate and rapid intelligent early warning of the early stage of the forest fire is realized, and the intelligent level and emergency response efficiency of forest fire prevention and control are significantly improved.