8705 results about "Network structure" patented technology

A method for detecting an infrared ship target based on an improved YOLOv7 is provided, including the following steps: obtaining an infrared maritime ship data set; reforming a YOLOv7 network structure based on an MobileNetv3 network and a bidirectional weighted feature pyramid network, and obtaining an infrared ship target detection model by introducing an attention mechanism and an optimized loss function; training and verifying the infrared ship target detection model based on the infrared maritime ship data set to obtain the infrared ship target detection model trained; and detecting a maritime ship based on the infrared ship target detection model trained.

The invention provides a power distribution network communication network fault diagnosis and self-adaptive reconstruction method, which comprises the steps of formulating corresponding layering and partitioning strategies aiming at different network scales and service types, including modular management of a large-scale multi-layer network and centralized response of a small-scale network; the faults of various networks can be effectively positioned and isolated; recognizing a reconstruction object after the fault positioning is completed, and formulating a network reconstruction scheme according to the recognized reconstruction object and the current state of the network, including but not limited to replacing a fault node, reconstructing a connection or modifying a route by starting a standby resource; after the reconstruction scheme is executed, the data collection and analysis period is adjusted according to the fault frequency, and the reconstruction interval is controlled to reduce redundant network structure adjustment and potential network fluctuation, including active detection and passive monitoring of the power distribution network and bidirectional monitoring of the network condition.

The invention provides a reinforcement learning-based database adaptive data flow acquisition optimization method and system, and relates to the technical field of databases, and the method comprises the steps of constructing a distributed acquisition proxy network, obtaining database data flow information, and carrying out multi-modal feature decomposition and fusion. And calculating an optimal initial acquisition parameter combination by using the fusion feature vector and combining a genetic algorithm, constructing a double-layer deep Q learning network structure, performing hierarchical training, and generating a hierarchical acquisition strategy model. The model adaptively adjusts a global acquisition strategy and parameters according to a characteristic drift degree and a performance index so as to realize continuous optimization of the acquisition strategy. According to the method, through multi-modal feature fusion, reinforcement learning and a self-adaptive tuning mechanism, the data flow collection efficiency is effectively improved, the collection delay is reduced, and the resource utilization rate is optimized.

The invention belongs to the technical field of charging station planning, and discloses a charging station planning method and system based on an automobile charging demand, and the method comprises the steps: building a space-time charging demand distribution map through multi-source data collection and fusion processing, carrying out the seasonal change analysis and future demand prediction, and obtaining a dynamic charging demand prediction model; site layout optimization under a multi-constraint condition is performed by combining an urban road network structure and traffic flow data to form a preliminary charging station layout scheme, and power grid load capacity evaluation and renewable energy access analysis are implemented to establish an energy collaborative supply guarantee system. And designing a peak-valley period charging price dynamic adjustment and appointment queuing mechanism to form an intelligent scheduling control strategy, and finally obtaining a diversified charging facility configuration scheme through different charging power requirements and vehicle type suitability evaluation. According to the method, the problems of inaccurate demand prediction, unreasonable layout, uneven resource allocation and the like in traditional planning are solved, and accurate matching of charging resources and user demands is realized.

The invention provides a network topology intelligent generation method and system based on deep learning and topology analysis, and relates to the technical network intelligence field, and the method comprises the steps: obtaining historical topology data, carrying out the feature extraction based on time sequence division, constructing a graph neural network model, training, and evaluating the evolution trend and stability of a network topology structure. And constructing a deep reinforcement learning model to generate an optimization strategy, and carrying out iterative optimization until a network topology structure meeting requirements is generated. The network structure can be adaptively optimized, the network performance is improved, the operation and maintenance cost is reduced, and the network stability is enhanced.

The invention discloses a drilling key parameter automatic regulation and control method based on an improved DDPG algorithm. The method comprises the following steps: collecting and preprocessing drilling data, and constructing a complete data set for training; a drilling speed (ROP) time sequence prediction model is constructed and trained on the basis of a self-attention mechanism (self-attention); building a simulation environment by using the trained ROP prediction model, and defining a state space, an action space and a multi-target reward function of a reinforcement learning DDPG algorithm model; a deep deterministic strategy gradient algorithm (DDPG) is adopted to train a reinforcement learning model, through an Actor-Critic network structure, an intelligent agent learns an optimal bit pressure (WOB) and rotating speed (RPM) regulation strategy in a simulation environment, and dynamic optimization of the drilling speed (ROP) is achieved. According to the method, through automatic and intelligent parameter regulation and control, the efficiency and stability of the drilling process are improved, the drill bit wear rate and the operation cost are reduced, and the method has good adaptability and multi-target optimization capacity and is suitable for drilling operation under the complex stratum condition.

A three-dimensional point cloud semantic segmentation method based on graph convolution and a grouping vector attention mechanism adopts a network structure of a U-Net style, feature extraction and propagation are performed on point cloud through multi-level feature coding and decoding, each level firstly centralizes and standardizes local neighborhood features through a geometric affine module, and then the local neighborhood features are subjected to feature extraction and propagation; the capturing capability of the model on the diversity of the local region is enhanced; secondly, constructing a local neighborhood graph, and designing an asymmetric edge feature learning function to extract local topological information; deep edge features are further extracted in combination with a residual network; and finally, selectively fusing local features and global context information under the condition of low parameter quantity by using a grouping vector attention mechanism, and performing cross-level feature interpolation up-sampling to an original scale to complete point cloud segmentation. The method can significantly improve the point cloud segmentation precision, and is suitable for the perception of a mobile robot and other devices carrying a three-dimensional sensor on a complex scene.

The invention relates to the technical field of computer vision and multi-modal learning, in particular to a weak supervision target detection method guided by cross-modal pseudo labels. According to the method, a labeled source domain data set is constructed to train an image classification teacher model, and a teacher-student network structure is constructed; clustering the regional features of the target domain image, allocating pseudo tags to each cluster by optimizing the allocation cost between the source domain category and the target domain cluster, and constructing a pseudo tag pool; and training a student model on the pseudo label pool for region feature detection of the target domain image. According to the method, a cross-modal attention mechanism is introduced, so that more accurate semantic alignment between a source category label and a target domain feature is realized; the stability of label distribution is improved by a structure keeping regular term; the generalization ability of the model is further enhanced by multiple rounds of pseudo-label confidence learning. The method can be widely applied to tasks such as target detection, cross-domain transfer learning and open world recognition, and efficient and accurate weak supervision target detection is realized.

The invention relates to the technical field of intelligent fault diagnosis, and discloses a fault diagnosis method for new energy power generation equipment, and the method comprises the steps: obtaining a multi-physical-quantity time dynamic data set; outlier elimination is carried out on the data set, and a deviation degree sequence of each physical quantity is extracted; based on the deviation degree sequence, normalization processing is completed, and a correlation matrix among multiple physical quantities is constructed; generating a preliminary network structure through correlation screening and symmetry completion; calculating a path weight based on the initial network structure and performing topology reconstruction to obtain a final network topology structure; performing deviation propagation analysis according to the final network topology structure, and determining potential fault position distribution; based on the fault position distribution, performing fault grade classification by adopting a support vector machine algorithm, and outputting a fault risk grade label; and in combination with the fault level label and the network topology structure, node risk scoring and area identification are executed, and finally fault positioning is completed. According to the method, accurate positioning of the fault position in a complex system can be realized.

The invention relates to the technical field of communication, in particular to a micro-segmentation-based multi-tenant zero-trust security system, which comprises a tenant network topology sensing module for acquiring network structures and service dependencies of tenants and generating a tenant-level network topology graph and a communication graph; a strategy rule automatic generation module generates a micro-segment access control strategy; the context-aware risk assessment module collects a user identity, an equipment state and a behavior track, and generates an access risk score; the strategy dynamic adjustment and application module updates the access control strategy; and the tenant isolation and zero-trust interaction module deploys a security sandbox to complete trust verification and isolation protection of cross-tenant requests. According to the method, the micro-segmentation strategy is generated by automatically identifying the service dependency boundary, the access control rule is dynamically adjusted based on the context, and refined isolation and risk-driven defense in a multi-tenant environment are realized in combination with the security sandbox and the zero-trust mechanism, so that the security and controllability of the system are remarkably improved.

The invention provides a coal mining digital twinning method based on multi-level dynamic deviation correction and AI optimization. The method comprises the steps of monitoring data acquisition and dynamic preprocessing, and performing denoising and space consistency detection on data; constructing and pre-processing the coal mine digital twinborn body; generating a nonlinear deviation field by adopting a high nonlinear variable interpolation and model mapping method; performing Kalman filtering error analysis and correction under the continuous analysis step, and marking an optimization model state in combination with a space mutation step; performing staged sliding window long-time-sequence fluctuation analysis, and classifying and marking regions according to an absolute value of an error and fluctuation; performing multi-level weighted error optimization based on a genetic algorithm, and adjusting material parameters and boundary conditions; carrying out material constitutive rectification based on dynamic Bayesian, and gradually enhancing the adaptability and accuracy of the model; and constructing a hybrid network structure AI optimization prediction model based on deep data driving, capturing the dynamic change of the material constitutive and the characteristic difference of the model constitutive, and realizing the precise digital twinborn simulation of the whole coal mining process.

The invention discloses a lightweight mobile phone screen defect detection method based on improved YOLOv10, and the method comprises the following steps: S1, collecting mobile phone screen image data, carrying out the enhancement processing of a data set, marking the data set, and constructing a mobile phone screen defect training data set; s2, adopting a Starnet module, a C2f-ODConv module and a DetectLSCD module to improve a YOLOv10 network model structure, and constructing a mobile phone screen defect detection model of a YOLO-STAR network structure; s3, using the mobile phone screen defect training data set constructed in the S1 as input, and training the improved YOLOv10 network structure model; and S4, deploying the trained YOLOv10 network structure model to a mobile device, and carrying out real-time detection on a mobile phone screen image. According to the invention, by improving the architecture of the YOLOv10 model, the complexity and the calculation amount of the model are reduced, the training data set and the detection process are optimized, and the deployment efficiency and the real-time detection performance of the model on equipment are improved.

The invention relates to the technical field of exception prediction, in particular to a storage cabinet exception trend prediction system based on time series data analysis, which comprises a state monitoring module, an interval sensing module, a path reconstruction module, a symptom activation module and an evolution prediction module. According to the method, the state vectors including the temperature, the voltage, the current and the door lock state are constructed and combined with the timestamp information to form the time sequence data sequence, and the dynamic expression mode of state change is established; a jump characteristic is analyzed by using a ratio of a time interval to a state change amplitude, a short-time disturbance path and a trend evolution path are distinguished by combining a jump rate statistical index, and an evolution activation signal is identified based on trend maintenance and non-fallback characteristics. On the basis, a neural network structure with long-time dependent learning ability is introduced to capture an aperiodic thermal anomaly trend in a state sequence, and the accuracy and timeliness of anomaly recognition are improved through multi-dimensional parameter cooperative processing and path construction logic.

The invention discloses an image super-resolution reconstruction method based on deep learning, and belongs to the technical field of image signal processing. Comprising the following steps: S1, realizing multi-dimensional feature representation of a low-resolution input image by using a deep learning network; s2, realizing accurate alignment and lossless mapping of cross-scale features, and effectively solving the problems of feature dimension inconsistency and information distortion; s3, explicitly learning an increment mapping relation between low-resolution input and a reconstructed high-resolution image, and relieving the gradient disappearance problem of a deep learning network structure; s4, enhancing the naturalness and detail fidelity of the reconstructed image through adversarial training; s5, integrating pixel-level reconstruction loss, perception loss and structural similarity loss, and effectively balancing objective indexes and subjective visual quality of the reconstructed image; and S6, realizing joint optimization of deep learning network parameters through a unified gradient back propagation mechanism, and finally outputting a high-quality super-resolution reconstructed image.

The invention provides an unstructured data automatic processing method and system based on deep learning, and relates to the technical field of data processing, and the method comprises the steps: obtaining unstructured data through an intelligent data collector, building a data feature fingerprint database, constructing a data quality evaluation model through an incremental learning algorithm, and carrying out the data standardization processing; extracting dominant features and recessive features by using a multi-layer cascade network structure and a self-attention mechanism, and constructing a feature association graph and a bidirectional reasoning channel; and performing rule matching and causal reasoning based on the domain knowledge base, and finally realizing automatic structuring of the unstructured data. According to the method, the automation degree and the accuracy of data processing are improved, the manual intervention cost is reduced, and the method has relatively high universality and expandability.

The invention provides a power distribution network system fault monitoring method and system based on multi-source information, and the method comprises the steps: obtaining a multi-source monitoring data set of a power distribution network system, carrying out the data feature extraction of the multi-source monitoring data set, and obtaining the time sequence change feature of operation state data and the space correlation feature of environment influence data; calling a pre-constructed fault detection model to carry out cooperative fault analysis on the time sequence change characteristics and the space correlation characteristics, and generating a fault detection result of the power distribution network system; determining the fault type in the power distribution network system and the position characteristic of the fault in the network structure according to the fault detection result; and generating a fault early warning instruction containing the fault positioning coordinate based on the fault type and the position feature. The overall practicability and reliability of power distribution network system fault monitoring are effectively improved.

The invention provides a power system optimization control method and device based on power market transaction. The method comprises the following steps: acquiring real-time electricity price fluctuation data, load demand prediction data and dynamic topology connection data between power grid nodes; generating a simplified network structure adaptive to a transaction scene based on the data, and adjusting an equivalent impedance parameter of a key transmission path in real time; a node voltage amplitude and a phase angle are cooperatively regulated and controlled through a multi-stage regulation mechanism of the solid-state transformer, and a power distribution scheme is formed; monitoring operation data of the solid-state transformer in real time by utilizing an edge computing node, and generating a local control instruction by combining the current equivalent impedance parameter; and carrying out space-time mapping matching on the local instruction and the electricity price fluctuation data, generating a combined control strategy through collaborative priority ranking, and dynamically adjusting a power distribution scheme. According to the technical scheme provided by the invention, collaborative optimization of economical operation and physical transmission performance of the power system is realized.

The invention relates to the technical field of intelligent logistics scheduling, in particular to a logistics scheduling planning method and system based on a graph neural network and reinforcement learning, and the method comprises the following steps: S1, constructing a dynamic graph structure of a logistics network; s2, carrying out embedded learning on the dynamic graph structure through a graph attention network, and extracting a multi-dimensional feature vector of each node; s3, inputting the multi-dimensional feature vector into a multi-agent reinforcement learning framework to generate an initial vehicle path planning scheme; s4, dynamically correcting the road section traffic state in the initial vehicle path planning scheme; s5, iteratively updating the vehicle path planning scheme through local reinforcement learning; and S6, outputting a final collaborative optimization cargo transportation track and a vehicle driving path. According to the method, dynamic modeling and multi-agent path collaborative optimization of a logistics network structure can be realized, and the method has adaptive adjustment capability on real-time traffic and environment change, so that the overall scheduling efficiency is improved.

The invention relates to the technical field of industrial chain monitoring and treatment, in particular to a monitoring method and system for industrial chain breakpoint treatment. The method comprises the steps that production, logistics, finance, policy and environment dynamic information is acquired through multi-source data, industrial chain comprehensive characteristics are generated through standardization, fractal dimension embedding expression and cross-dimension fusion, and historical trend dependency is introduced to enhance prospective prediction; a dynamic coupling network is constructed, and risk propagation intensity between nodes is quantified by using a dynamic edge weight, so that cross-level breakpoint propagation analysis is realized; risk indexes are calculated by fusing node features and a network structure, breakpoint candidate nodes are screened in combination with an adaptive threshold value, and a multi-step evolution trend is predicted by adopting a nonlinear propagation function and mapped into a multi-level early warning level. And generating a governance strategy according to the risk level, and evaluating the effect in real time and dynamically adjusting parameters through a closed-loop optimization mechanism. According to the invention, closed-loop management of risk identification, prediction and adaptive treatment is realized.

The invention discloses an urban life body full-cycle monitoring system based on a digital twinborn technology, relates to the technical field of urban governance, integrates multi-source data such as GIS data, remote sensing information, surface permeability data and an underground drainage pipe network structure, quantifies a flood high-risk area, and provides an urban life body full-cycle monitoring system based on the digital twinborn technology. The AI flood prediction module deduces the ponding range and the drainage capacity under different rainfall situations by adopting extreme weather analogue simulation based on a historical rainfall trend, a current meteorological condition and a drainage network state, and dynamically adjusts a flood diffusion trend prediction model, so that the flood risk identification is more accurate; the intelligent emergency scheduling module automatically matches flood control resources, combines flood high-risk areas, ponding point locations and traffic flow analysis, optimizes a drainage pump station start-stop strategy, remotely regulates and controls an underground drainage gate, dynamically adjusts the overflow adjustment capability of a sewage treatment plant, and optimizes an emergency risk avoiding route in combination with an intelligent traffic management system. And the accuracy and the execution efficiency of the flood control scheduling scheme are improved.

A fabric control protocol (FCP) is a data transmission protocol that enables spraying of individual packets for a given packet flow across a data center from an ingress interface of the source data processing unit (DPU) across a plurality of parallel data paths of a logical tunnel in the network fabric to the egress interface of the destination DPU. The FCP has congestion control mechanisms used to determine a degree of congestion at the egress interface of the destination DPU and to modify a send window size at the source DPU based on the degree of congestion. Reliable FCP (rFCP) extensions provide reliability enhancements and improved failure resilience within the data center. The rFCP extensions provide an unsolicited mode for low latency operation with enhanced reliability mechanisms. The rFCP extensions provide failure resilience mechanisms to identify and avoid failed paths among multiple parallel data paths within the logical tunnel.

Disclosed is a method for establishing a 3D medical image segmentation model based on masked modeling and application thereof includes: establishing a semi-supervised learning network, wherein a student network includes an encoding module for extracting latent features and a segmentation decoder that predicts segmentation results, a teacher network includes an encoding module and a segmentation decoder that are structurally consistent with the student network; training the semi-supervised learning network, wherein during training, two random masking operations are performed on each image, and the image is input to the two networks respectively; optimizing and updating the weight of the student network, and transferring the updated weight to the teacher network, wherein the training loss function includes prototype representation loss, which is used to characterize the difference between the prototypes extracted and generated by the two networks; the student network may further include a reconstruction decoder and an auxiliary segmentation decoder.

The invention discloses an AI model adaptive tuning method based on deep reinforcement learning, and belongs to the technical field of industrial process control, the method comprises the following steps: firstly, constructing a deep reinforcement learning environment comprising a state space, an action space and a reward function, the state space comprising a process variable measurement value, a controller parameter and a performance index; the action space is a parameter adjustment quantity vector, and the reward function comprehensively considers the stability, rapidity and accuracy of the system; then, based on a deep Q learning network of a dual-network structure, training data are stored and sampled by adopting an experience playback mechanism, and network parameters are iteratively optimized through a deep Q learning algorithm; and finally, realizing online adaptive adjustment of controller parameters by using the optimized deep Q network. The method has the advantages of being high in adaptability, good in optimization performance, high in practicability and the like, intelligent optimization control over an industrial process control system can be achieved, and the overall performance of the system is improved.

The invention discloses a knowledge distillation method, electronic equipment and a computer readable storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: carrying out the grouping of an attention map matrix of a teacher network based on the number of attention heads in an initial network structure of a student network, a high-dimensional attention space of a teacher network can be divided into subspaces with the same dimension as a student network, and attention map matrixes in groups are spliced, so that each group of spliced matrixes is aligned with an attention head of an initial network structure, a relatively regular corresponding relationship between the attention maps of the teacher network and the student network is realized, and the attention map splicing efficiency is improved. Therefore, the distillation loss can be calculated, and the knowledge of the complex teacher network can be accurately transmitted to the lightweight student network. Therefore, the problem that in a multi-head attention mechanism, the number difference between the teacher network and the student network hinders the student network to fully learn the teacher network knowledge can be solved, and the technical effects of improving the knowledge distillation effect and enhancing the student network performance are achieved.

The invention discloses a photovoltaic panel surface defect detection method based on a pruning multi-scale feature fusion network, and belongs to the technical field of photovoltaic panel defect detection, and the method comprises the following steps: S1, data set processing; s2, constructing a network; s3, model training; and S4, defect detection. According to the method, a feature fusion pruning efficient multi-scale attention (FFPEMA) module is introduced into a backbone network to improve the feature representation capability, and network structure global pruning is combined with a feature fusion technology to enhance feature representation so as to improve the model detection precision; a dilated convolution content awareness (DCCARAFE) module is designed, a dilated convolution strategy is adopted, the receptive field is increased by adjusting the voidage, and the flexibility of model feature extraction is improved; a pruning efficient self-adaptive feature integration fusion strategy (PAFIFN) is provided, and the feature extraction and learning ability of the model for defects of different sizes and scales is improved.

The invention discloses an energy equipment carbon emission scheduling optimization method in an integrated energy system, and the method comprises the steps: determining an energy equipment carbon emission factor, constructing a carbon emission cost function, employing Monte Carlo simulation to generate a carbon transaction price fluctuation scene, and carrying out the carbon transaction price fluctuation scene. The method comprises the following steps: calculating a carbon transaction risk index of equipment in each scene according to a cost function, determining start-stop time by economic benefit optimization, constructing a comprehensive energy system topology network structure, updating a carbon flow path set according to the start-stop time, calculating a comprehensive carbon emission coefficient, establishing an energy flow and carbon flow coupling model between a high emission path and a low emission path, and constructing a comprehensive energy system topology network structure. A scheduling strategy feasible space is obtained through a cooperation strategy, and an optimal cooperation scheduling operation strategy is obtained through optimization of a particle swarm algorithm by taking the minimum carbon emission as a target. The method can effectively reduce the carbon emission / optimize the start and stop of the equipment, improves the economical efficiency and environmental protection property of system operation, achieves the dual goals of energy conservation and emission reduction and benefit improvement, and has good interpretability at the same time.

The invention discloses an ecological restoration quantification method based on a watershed ecological safety complex network. The method comprises the following steps: S1, obtaining multi-time-sequence watershed multi-source heterogeneous spatial data of a target area; s2, taking the grid as an evaluation unit, establishing a drainage basin ecosystem service function evaluation system, and obtaining spatial-temporal change characteristics of an ecological safety function; s3, analyzing the spatial-temporal change characteristics of the ecological land scale of the watershed; s4, constructing a drainage basin ecological resistance surface; s5, identifying an ecological corridor; s6, constructing an ecological safety complex network model by taking the ecological source as a node, the ecological corridor as an edge, the scale of the ecological source as a point weight and the proportional relation between the connection efficiency of the ecological corridor and the corridor resistance as an edge weight, and extracting spatial-temporal evolution characteristics of an ecological safety structure; s7, a coupling coordination degree model is adopted to simulate and analyze the characteristic relation among the safety function, the safety structure and the land use scale, and the coupling coordination degree is obtained; and S8, guiding and optimizing an ecological restoration implementation strategy. According to the method, ecological security network structure features on the watershed system scale are quantified.

The invention discloses a stability analysis method of a network tracking / constructing hybrid station based on an impedance model, and relates to a modeling and stability evaluation technology of a wind power plant grid-connected system. The method comprises the following steps: firstly, respectively establishing sequence impedance models of a following network type direct-driven fan and a construction network type direct-driven fan based on a harmonic linearization method, and verifying the accuracy of the established models in different frequency bands by combining an impedance sweep frequency method; secondly, on the basis of considering the influence of line impedance, deducing to form an impedance network structure of a hybrid wind power plant grid-connected system, and constructing a stability judgment condition suitable for system analysis; and finally, based on the established impedance stability criterion, identifying possible oscillation risks of the system in subsynchronous and supersynchronous frequency bands. The method is suitable for wind power system stability analysis in various hybrid access scenes, and provides theoretical basis and technical support for subsequent system parameter setting, control strategy optimization and oscillation suppression measures.

The invention discloses a reinforcement learning method and system for source network load storage collaborative multi-scene optimization, and the method comprises the steps: constructing a power distribution network optimization model with the minimum cost, and converting a mixed integer nonlinear problem into a solvable mixed integer second-order cone optimization problem; converting a mixed integer second-order cone optimization problem into a reinforcement learning decision model, and performing multi-round assignment on boundary condition parameters by using different operation scene data to form a multi-scene training task set; designing a reinforcement learning decision model multi-scene training loss function, combining the reinforcement learning decision model to construct a strategy neural network, an evaluation network neural network and a scene representation embedded neural network, and completing reinforcement learning decision adaptive to multi-scene optimization; and training a power distribution network multi-scene optimization decision model based on the multi-scene training task set, the multi-scene training reinforcement learning loss function and the neural network structure, and deploying the power distribution network multi-scene optimization decision model to an actual system to complete reinforcement learning decision model application oriented to source network load storage collaborative multi-scene optimization.

The invention discloses a construction man-machine safety intelligent detection and early warning method based on deep learning. The method comprises the following steps: S1, introducing a wavelet convolution improved YOLOv11 target detection algorithm framework; s2, coordinate extraction of the instance segmentation module; s3, designing a monocular depth estimation network; s4, training the monocular deep network; and S5, converting the two-dimensional coordinates into the three-dimensional coordinates. According to the invention, the real distance between the worker and the equipment is obtained and real-time safety early warning is carried out; obtaining a man-machine actual distance, and judging the accurate position and posture of the machine by combining angle information of the construction machine with edge contour detection; target detection in a complex construction scene is realized by optimizing a network structure; the S2R-DepthNet depth estimation network is used for performing depth estimation on a single RGB image through a self-supervised learning method, the three-dimensional information acquisition process is simplified, and dependence on expensive hardware is eliminated.