503 results about "Short-term memory" patented technology

Disclosed is a suppression method for a multipath signal of an image mode based on correlation peaks of a satellite baseband signal, including the steps of: constructing a real-world direct-multipath two-dimensional color image mode data set; and building a multipath suppression model of a deep learning network based on a long-short term memory (LSTM) and a self-attention mechanism module, and training the model. In the present disclosure, a satellite signal can be quickly captured without losing the sensitivity of the captured signal, and a Beidou satellite baseband signal can be captured from a signal recorded in the real world, which enables the model to learn complex signal patterns, improving the accuracy and robustness of multipath signal suppression in urban complex scenes.

The invention relates to the technical field of environmental protection, in particular to a big data monitoring and intelligent early warning platform and method for environmental protection. Comprising a multi-source data acquisition module, a data integration and preprocessing module, an environmental risk factor identification module, a real-time monitoring and anomaly detection module, a prediction model and early warning module, a self-adaptive early warning and feedback module and a decision support and visualization module. The data integration and preprocessing module preprocesses multi-source data, the environmental risk factor identification module identifies key risk factors by using an XGBoost algorithm, and carries out accurate prediction and early warning on risks through a random forest and long and short term memory network combination model, and an adaptive feedback mechanism dynamically adjusts an early warning threshold value, so that the early warning efficiency is improved. According to the invention, the intelligent level and the scientific level of environmental protection are obviously improved.

The invention relates to the technical field of pumped storage power station multi-sensor monitoring network time-space synchronous measurement, in particular to a pumped storage power station multi-sensor safety monitoring system, which comprises a sensor data acquisition module, a time-space synchronous network module, a data calibration module, a dynamic association model construction module, a multi-dimensional data fusion module and a graded early warning module. The dynamic correlation model building module simulates dynamic response of seepage pressure and surrounding rock strain through a fluid-structure interaction equation and a discrete element digital twinborn model, updates fracture permeability boundary conditions in real time, and extracts cross-sensor space-time correlation characteristics by adopting a convolutional long-short-term memory network improved by a multi-head attention mechanism. According to the method, the real-time early warning precision of abnormal events and the reliability of structural health assessment in the complex hydraulic and mechanical coupling environment are improved.

The invention relates to the technical field of automatic logistics, in particular to an intelligent warehouse cargo scheduling device, which comprises a dynamic priority adjustment module, a self-adaptive classified storage module, a cargo flow optimization module, a resource allocation module, a fault analysis and maintenance module, a behavior monitoring module, an energy management module and a cost control module. According to the method, the cargo scheduling priority is optimized through a support vector machine algorithm and a decision tree method, the warehouse space utilization rate is improved by utilizing a convolutional neural network and a genetic algorithm, the cargo flow is optimized by adopting linear programming and a network flow analysis algorithm, and the cargo flow trend is estimated through an autoregressive integral moving average model and a recurrent neural network. The system utilizes a long short-term memory network to identify equipment faults and maintain, reduces the risk of accidental shutdown, improves the operation safety through a computer vision technology and a behavior identification algorithm, and utilizes an energy consumption mode analysis and cost-benefit analysis method to reduce the operation cost.

The invention discloses a CCER forestry carbon sink project development system based on artificial intelligence technology monitoring, and relates to the technical field of carbon sink monitoring and optimization. According to the CCER forestry carbon sink project development system based on artificial intelligence technology monitoring, space-time estimation is carried out on CCER forestry carbon sink project data through the deep learning model, and the dynamic change of the carbon sink amount is accurately predicted. Spatial features are extracted through a convolutional neural network, time sequence data are analyzed in combination with a long-short-term memory network, and a carbon sink prediction model is constructed. And through statistical analysis and a spatial interpolation algorithm, carbon sink quantity change is monitored in real time, an abnormal trend is identified, and early warning is provided. Based on a mixed integer linear programming algorithm, forest land resource allocation and vegetation types are optimized, a carbon sink optimization scheme is generated, and carbon sink benefit maximization is promoted. And through an intelligent carbon sink management scheme, sustainable ecological development and green and economic popularization are assisted.

The invention provides a method and a system for identifying and cleaning abnormal data, which relate to the technical field of data processing, and are used for identifying and cleaning abnormal data by acquiring real-time data flow, determining a self-adaptive threshold value by using a trained Gaussian mixture model and detecting abnormal data points. A deep belief network and long and short term memory network composite model is adopted to carry out deep feature learning and anomaly verification, and high-level feature representation and anomaly degree quantitative evaluation results are obtained; and evaluating a conditional dependency relationship among the abnormal data points by using the probability graph model, constructing a target probability dependency graph, and generating an abnormal data point recognition result. Based on an identification result, a reward mechanism excitation algorithm is applied to optimize a cleaning strategy; according to the method, the to-be-determined abnormal region is identified, the multi-modal fusion technology is introduced to perform correlation analysis, and the data in the to-be-determined abnormal region is judged and cleaned based on the analysis result to obtain the target cleaning strategy, so that the abnormal data identification and cleaning efficiency and accuracy are improved, and the method has relatively high adaptive capability and expansibility.

The invention provides an intelligent memory leak prediction and tracking method and system for a micro-service architecture, and relates to the technical field of memory leak prediction, and the method comprises the steps: collecting the runtime data of a service node, constructing a service calling topological graph, extracting the time sequence and topological features, and carrying out the feature fusion. And constructing a deep learning prediction model by combining a bidirectional long-short term memory network with a multi-head attention mechanism and a graph attention network, and predicting a memory use trend and a risk score of a service node. Potential memory leak nodes are identified according to the risk score and the abnormal degree predicted value, a key propagation path is identified through a service calling topological graph and a minimum cut algorithm, and finally a root service of memory leak is positioned by using a belief propagation algorithm, so that accurate prediction and rapid positioning of the memory leak in the micro-service system are realized. And the system stability and reliability are improved.

The invention discloses an intelligent monitoring system for sewage treatment in the coal chemical industry, and relates to the technical field of sewage treatment.The intelligent monitoring system is characterized in that a water quality monitoring unit is used for collecting parameters such as COD, turbidity, pH and heavy metal ion concentration in real time, the change trend of water quality in a short time is predicted through a long-short term memory (LSTM) network, and then adding parameters of a flocculating agent are dynamically calculated; the defects of a fixed adding curve or artificial experience adjustment are overcome, so that the adding amount, the type and the speed of the flocculating agent can be adaptively adjusted to be matched with the change of sewage components; the addition optimization unit is combined with a CNN-LSTM model to analyze the rheological characteristics of the sludge, so that the addition of the flocculating agent adapts to the rheological change of the sludge, the flocculation effect is optimized, and the sludge settling performance is improved; meanwhile, the stirring control unit adjusts the stirring strength and time in real time based on ultrasonic online monitoring data, so that the flocculating agent is fully dispersed, the particle distribution state is optimized, the flocculation reaction is more uniform, and the subsequent precipitation separation efficiency is improved.

The invention relates to the technical field of e-commerce intelligent recommendation, in particular to an e-commerce intelligent recommendation system based on user behavior big data, which comprises a data acquisition and storage unit, a data preprocessing unit, a model construction training unit and a recommendation generation and output unit. The method comprises the steps that data such as user behaviors, equipment, geographic positions and access time are collected and stored in a distributed database, a data preprocessing unit extracts the data, constructs a user behavior sequence and encodes commodity and behavior characteristics, and a model construction training unit obtains user interest vectors by using a long and short-term memory network in combination with an attention mechanism; model parameters are updated with the help of a deep Q network, commodities are screened, sorted and displayed by a recommendation generation and output unit according to interest vectors, meanwhile, a feedback optimization system is collected, recommendation accuracy is improved, user shopping experience is enhanced, and an e-commerce platform is assisted to improve operation efficiency and commodity sales conversion rate.

The invention relates to the technical field of artificial intelligence, and discloses an artificial intelligence-assisted adaptive sickbed control system and method, and the method comprises the steps: carrying out the noise reduction and normalization preprocessing of original physiological data collected by a multi-modal sensor, and obtaining standardized multi-modal data; performing time sequence fusion and dimension reduction to obtain low-dimensional fusion features; performing feature extraction through a multi-branch convolutional neural network to obtain generalization feature representation; the optimal posture demand of the patient is predicted through a long-short term memory network model with an attention mechanism, and an optimal posture demand prediction result is obtained; a layered sickbed adjustment strategy is generated through a double-time-scale reinforcement learning network comprising a priority experience playback mechanism; according to the intelligent sickbed control method and system, the multi-mode sensing data can be comprehensively utilized, the adaptive learning ability is achieved, the intelligent sickbed can be accurately controlled, and the performance and the application effect of the intelligent sickbed are improved.

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

The invention provides an intelligent refrigeration system efficiency optimization method, system and device based on deep learning, and relates to the technical field of intelligent refrigeration system efficiency optimization. The method comprises the following steps: firstly, collecting historical operation parameters and environmental data of the refrigeration system, preprocessing, and integrating into a sample set according to a timestamp; and based on the hybrid model of the convolutional neural network and the long-short term memory network, constructing a state prediction model for predicting the system operation state. And in combination with the state prediction model, an optimization objective function is defined for data prediction at the next future moment, and an optimal adjustment strategy for system operation is formulated by using a deep Q network model. And finally, monitoring current operation parameters and environment data in real time, calculating a comprehensive safety score, comparing the comprehensive safety score with a safety threshold value, and deciding whether to continue to optimize the efficiency or preferentially carry out troubleshooting. According to the method, through combination of deep learning and physical modeling, efficient and safe refrigeration system operation optimization is realized.

The invention discloses a building crack early prediction and restoration method based on a machine learning algorithm, and the method comprises the following steps: collecting stress, displacement, temperature and image data through a self-adaptive sensor network, and enabling a sensor to dynamically adjust the collection frequency and position; preprocessing the collected multi-source data, including de-noising, data completion and image enhancement, and extracting crack features; performing cross-scale multi-dimensional feature extraction based on the preprocessed data, and generating high-dimensional features for crack prediction; inputting the high-dimensional features into a fusion quantum Bose-Einstein condensation model and a long-short term memory network model, and performing dynamic space-time prediction of the crack; according to the prediction result, generating a repair suggestion by using an intelligent optimization algorithm; after the repair plan is implemented, the repair effect is continuously monitored through the sensor network; the sensor network configuration is adjusted in real time, and the monitoring and prediction accuracy is improved. According to the invention, early prediction and effective repair of the building crack are realized.

The invention discloses a multi-channel gas analysis fire-fighting facility fire early warning system, and belongs to the technical field of artificial intelligence. According to the invention, corresponding floors and corresponding equipment in a building are positioned, fire-fighting equipment information, hospital fire records and equipment alarm records are obtained, and equipment states are analyzed; the method comprises the following steps: monitoring data in a building in real time, unifying a time axis, forming fire early warning data, training a long and short-term memory network model, analyzing the fire early warning data, generating a fire early warning level, and positioning in a three-dimensional model of the building; acquiring real-time gas component data of a fire early warning position area, training a fire degree evaluation model, predicting newly acquired data, judging whether a fire exists in a current environment or not, evaluating the severity of the fire, identifying misinformation and adjusting an early warning level; and according to the confirmed early warning level, a corresponding alarm mechanism is triggered, fire fighting equipment is automatically linked, an evacuation instruction is generated under the condition of advanced early warning, and personnel evacuation guidance is carried out.

The invention relates to the technical field of subgrade settlement prediction, in particular to a subgrade settlement prediction method based on deep learning, electronic equipment and a medium, the subgrade settlement prediction method provided by the invention adopts a subgrade settlement prediction model for prediction, and the model adopts the combination of a convolutional neural network and a bidirectional long-short term memory network. Spatio-temporal features are effectively captured, and the prediction precision and stability of the model are improved; by introducing a channel attention enhancement mechanism module, dynamic weighting is performed on different feature channels, the selection ability of the model to key features is improved, and the sensitivity of the model to subtle changes in data and the capture ability of the model to global features are enhanced; the model is trained through the genetic algorithm, the efficiency is improved, and the genetic algorithm can automatically find an optimal combination in a complex hyper-parameter space, so that the model can show good generalization ability on different roadbed settlement data sets, and the prediction precision of the model is improved.

The embodiment of the invention provides an elderly emotion recognition method and device based on multi-modal perception, and the method and device achieve the optimization and enhancement of the signal quality through innovatively constructing a multi-modal data preprocessing mechanism and integrating the facial expression, voice and posture features. And designing a personalized feature mapping model based on historical emotion expression data, and establishing an adaptive feature fusion strategy for intelligent matching in combination with a cross-modal attention network. A hierarchical time sequence classification mechanism is introduced, dynamic modeling of the emotional development trend is realized through a long-short term memory network, and accurate prediction of the emotional state is supported. According to the method, the defects of the traditional technology in the aspects of multi-modal processing, personalized modeling, time sequence analysis and the like are effectively overcome, and the accuracy and reliability of sentiment recognition of the old people are remarkably improved.

The invention discloses a Multi-agent-based psychological counseling platform and method. The counseling platform comprises an evaluation agent used for dynamically extracting psychological features based on dialogue behaviors of a visitor agent and updating results to a traits library module; the planning agent is used for constructing a personalized psychological counseling scheme based on the traits library module and dynamically adjusting a counseling process and a strategy according to counseling feedback; the consultant agent is used for carrying out dialogue interaction with the visitor agent according to the personalized consultation scheme and implementing emotion pacification and psychological intervention; the visitor agent is used for expressing psychological troubles, feeding back psychological state changes and promoting the consultation process; the historical dialogue long and short-term memory module is used for storing multiple rounds of psychological counseling interaction contents and comprises historical dialogue short-term memory and historical dialogue long-term memory; and the dynamic probability memory retrieval module is used for generating probability distribution based on the current consultation context and dynamically calling memory nodes from the historical dialogue long and short term memory module or the traits library module.

The invention discloses a water turbine wear fault diagnosis method based on a deep learning model, and belongs to the technical field of water turbine fault signal processing. Comprising the following steps: acquiring a water turbine wear fault sound signal; a Tent chaotic mapping function is adopted to optimize an initial population of a white shark optimization algorithm WSO; introducing a long and short term memory (LSTM) network model to solve the limitation of a convolutional neural network (CNN) in processing a long sequence problem, and constructing a CNN-LSTM water turbine wear fault diagnosis model; an improved white shark optimization algorithm IWSO is introduced into the CNN-LSTM model, and hyper-parameters of the model are optimized, so that the diagnosis precision is improved, and the calculation amount is reduced; and importing the collected fault signal as an input into the model for training and testing to obtain a final diagnosis result. Diagnosis and identification of the water turbine wear fault signal are completed through the steps, the deep learning model has good robust performance, the diagnosis precision of the water turbine wear fault can be improved, and the deep learning model serves as a beneficial supplement of an existing water turbine state monitoring and fault diagnosis system.

The invention belongs to the technical field of mobile communication, and particularly relates to a multi-agent joint edge caching method based on a federated learning architecture, which is suitable for a macro-micro heterogeneous network environment. According to the method, the content popularity is predicted, and a cache decision is optimized by adopting a deep reinforcement learning method; according to the method, distributed prediction is carried out on content popularity through a federated learning architecture and a denoising auto-encoder long-short term memory network, and dynamic optimization of content cache is realized. And optimizing a caching strategy and resource allocation by combining a game model between the micro base station and the macro base station so as to maximize the effectiveness of the infrastructure provider. According to the method, a multi-agent deep reinforcement learning algorithm without information sharing is introduced to optimize the system efficiency. The system reduces transmission delay and improves network performance and user experience by calculating a dynamic adjustment strategy of terminal user content popularity and combining a communication resource allocation strategy. The method effectively solves the problems of high delay, low cache resource utilization rate and the like in the existing system, and is suitable for a complex network environment.

The invention discloses a sea surface chlorophyll concentration prediction method and system based on remote sensing and deep learning, and relates to the technical field of marine ecological environment, and the method comprises the steps: inputting the marine chlorophyll a data of a target region into a trained marine chlorophyll concentration prediction model, and obtaining the marine chlorophyll concentration information of the target region; wherein the prediction model is an improved marine chlorophyll a concentration prediction model based on a convolutional long short-term memory network; the periodic feature extraction module is used for extracting historical synchronous related periodic features based on historical synchronous chlorophyll concentration data, and the noise removal and fusion module is used for eliminating noise output by the recent spatial-temporal feature extraction module and the periodic feature extraction module and fusing output results to obtain a marine chlorophyll a concentration data prediction result. According to the invention, the sea surface chlorophyll concentration can be predicted more accurately.

The invention relates to a target trajectory prediction method and system based on a bidirectional long-short-term memory network, and the method mainly comprises the following steps: firstly, carrying out the spatial-temporal feature extraction of target trajectory data through a bidirectional long-short-term memory network layer, obtaining the spatial-temporal feature vector of target data, capturing a long-term dependency relationship and a short-term dependency relationship, and carrying out the spatial-temporal feature extraction of the target trajectory data; wherein the long-term dependency describes the time sequence and continuity of motion, and the short-term dependency describes the maneuvering characteristics of the motion; secondly, a prediction model is constructed, and hyper-parameters of the prediction model are optimized, so that the prediction performance and generalization ability of the model are improved; and finally, predicting a future trajectory of the target by using the trained optimization model, outputting a trajectory coordinate sequence of a plurality of time points in the future, and realizing accurate prediction of the target in a period of time in the future. According to the method, the bidirectional long-short-term memory network and hyper-parameter optimization are combined, so that high-precision trajectory prediction of the target is realized, and the adaptability of the prediction model to complex motion characteristics is enhanced.

The invention provides a vehicle track prediction method and device suitable for different driving styles, and relates to the technical field of automatic driving. The vehicle trajectory prediction method suitable for different driving styles comprises the following steps: acquiring a sample data set of an original vehicle trajectory; filtering the sample data set to obtain a filtered sample data set; performing dimension reduction processing by adopting a principal component analysis method to obtain dimension-reduced data; carrying out soft clustering processing by adopting a fuzzy C-means clustering algorithm to obtain a hard tag; constructing an initial global-local attention mechanism-based residual bidirectional long-short term memory network; training the initial network to obtain a trained bidirectional long-short term memory network based on the global-local attention mechanism residual error; acquiring state information of the target vehicle in the observation time domain; and inputting the state information into the trained network to obtain the future trajectory of the target vehicle in the prediction time domain. According to the invention, the accuracy of vehicle trajectory prediction can be improved.

The invention provides a real-time network security monitoring and protection method and system based on deep learning, and the method comprises the steps: extracting the local features of network traffic through a convolutional neural network, analyzing the time sequence features of abnormal network traffic through a long-short term memory network, and recognizing a network attack type; calculating the intensity of the network attack based on the traffic rate, the duration and the number of source IPs of the network attack, calculating a risk score based on the local features of the network traffic, and carrying out weighted calculation on the risk score and the intensity of the network attack to obtain a comprehensive score; and in response to different types of network attacks and in combination with the comprehensive scores of the different types of network attacks, executing different network attack protection measures, adjusting the protection level in real time according to the strength and risk scores of the network attacks, recording network attack information and protection measures, and generating a security log. The network flow can be analyzed in real time, the network attack type can be identified, corresponding protection measures can be taken, the accuracy of network attack detection is improved, and the risk is reduced.

The invention relates to the technical field of data center energy control, and discloses a data center cooling system dynamic control method adaptive to equipment aging, and the method comprises the steps: firstly constructing an energy consumption model for controlling a data center cooling system based on equipment parameters and an energy conservation principle; in order to solve the problem of performance degradation possibly caused by long-term operation of equipment, the method monitors operation parameters of key equipment such as a cooling tower, a heat exchanger, a water chilling unit and an air conditioner terminal in real time; according to the method, an equipment energy consumption model is iterated in a rolling manner by using a long-short term memory (LSTM) network model in deep learning, so that the energy consumption model for control is dynamically updated, the prediction precision is improved, and finally efficient model prediction control is realized. The method is not only suitable for the design and construction of a new data center, but also suitable for the transformation and upgrading of an existing data center, and has wide application prospects and practical values.

The invention relates to the technical field of intelligent control, in particular to a limo equipment wireless intelligent control method and system based on AI and voice recognition. The method comprises the following steps: a user sends a voice command, denoising and enhancing voice signals through a beam forming technology and an adaptive filter, enhancing a time-frequency structure of the voice signals by using Fourier transform and dynamic gain, and performing auditing verification by generating an auditing code; extracting multi-modal features, generating high-dimensional hidden representation through a bidirectional long-short term memory network and an attention mechanism after fusion, optimizing text output, and improving text quality by using spelling error correction and semantic consistency verification; then analyzing the instruction and generating a standardized instruction format through an intention recognition and slot filling model, executing rule verification to ensure the reasonability and safety of the instruction, and finally generating a control command in combination with environmental data to ensure the intelligent control of equipment. According to the invention, efficient and intelligent voice control and safety verification of the motor home equipment are realized.

The invention provides a metallurgical process optimization method and system for a low-oxygen low-nitrogen aluminum-vanadium alloy, and relates to the technical field of process optimizing.The method comprises the steps that correlation characteristics among process parameters are extracted through a three-layer gating map attention network, and a parameter coupling state vector is constructed in combination with a time sequence attention module; constructing a process prediction model based on the state vector, and fusing a thermodynamic equilibrium equation and a long-short term memory network to predict the oxygen and nitrogen content; a deep reinforcement learning network of a Soft Actor-Critic algorithm is adopted, the parameter coupling state vector and the oxygen and nitrogen content prediction value serve as state input, and the technological parameter adjustment amount is output; and planning and iteratively training a future adjustment strategy through Monte Carlo tree search to finally obtain an optimal process parameter combination meeting oxygen and nitrogen content constraints, thereby realizing intelligent optimization control of the metallurgical process.

The invention belongs to the technical field of ship fuel consumption prediction and big data analysis, and relates to an operating ship fuel consumption prediction method based on a deep learning model. The method comprises the following steps: 1, synchronously acquiring multi-source data from a ship automatic identification system, a cabin monitoring system and an ERA5 meteorological database, and preprocessing the multi-source data; step 2, inputting the preprocessed data into a pre-trained CNN-BiLSTM-Attention model to carry out ship fuel consumption prediction; wherein the CNN-BiLSTM-Attention model is composed of a convolutional neural network, a bidirectional long and short term memory network and a time attention module; the model adopts a composite loss function based on navigational speed-power-fuel oil physical constraint. The method is more excellent in the aspects of prediction precision, robustness and fitting effect. Compared with the prior art, the method has higher accuracy and reliability in the aspect of ship fuel consumption prediction.

The invention discloses a dam effect quantity prediction method and system, a medium and equipment, and relates to the technical field of safety engineering, dam monitoring data is acquired, influence factors of a dam effect quantity are determined, and effect quantity influence factor samples are constructed; performing dimension reduction processing on the influence factor sample through a principal component analysis method to obtain an influence factor matrix after decoupling and dimension reduction; coupling the long and short-term memory neural network with an artificial fish swarm algorithm, constructing a dam effect quantity prediction model, performing parameter setting on the dam effect quantity prediction model, and optimizing the initial weight and the number of neurons of the long and short-term memory neural network by using the artificial fish swarm algorithm; training the optimized long and short term memory neural network until convergence, and obtaining a trained dam effect quantity prediction model; and performing prediction and anti-normalization processing to obtain an effect quantity prediction result. And the accuracy, the stability and the efficiency of dam effect quantity prediction are improved.

The invention provides a medical data anomaly detection and automatic acquisition method, and aims to overcome the defects of single anomaly detection accuracy, response speed and data acquisition strategy in the prior art. According to the method, time-frequency characteristics of medical data are extracted through wavelet transform, and a self-adaptive multi-scale anomaly detection model is constructed by using a long-short-term memory network, so that a time sequence dependency relationship of the data is learned and parameters are dynamically adjusted; the intelligent acquisition strategy optimizer automatically adjusts an acquisition strategy based on a reinforcement learning technology to adapt to new medical data changes. According to the invention, the accuracy and real-time performance of anomaly detection are improved, the data acquisition strategy is optimized, and the intelligent level of medical data management is enhanced.

The invention discloses a heat supply load prediction system, and relates to the technical field of heat supply prediction. The data acquisition and fusion module is used for acquiring multi-source heterogeneous data in real time and integrating the acquired multi-source heterogeneous data into a data set with a uniform format; the feature engineering module extracts key features influencing the heat supply load, and a feature subset with the highest prediction value is screened out through a feature selection method and serves as input of a deep learning model; constructing a deep long-short-term memory network model based on an attention mechanism, and performing model training through the training data set; the real-time prediction and adaptive adjustment module receives new data acquired by the data acquisition and fusion module in real time, and inputs the new data into the trained deep long and short term memory network model based on the attention mechanism for real-time prediction of the heat supply load; by integrating multi-source heterogeneous data and utilizing an advanced deep learning model, accurate prediction of the heat supply load is realized, and the method has real-time learning and adaptive adjustment capabilities.