592results about How to "Increase reach" patented technology

The invention relates to the technical field of business adaptation, in particular to a business adaptation degree intelligent analysis method and system combined with a customer portrait. The method comprises the following steps: firstly, collecting customer data from a plurality of data sources, then processing the collected data, extracting interest labels of customers, constructing a customer-interest label matrix, comprehensively evaluating the purchase ability and loyalty of the customers, updating the portraits of the customers according to the interest preferences of similar customers, and then collecting the business demand data of the customers. The method comprises the following steps: constructing a business demand model, quantifying the demand intensity of a customer, generating a demand score vector, matching a customer portrait with the demand score vector, calculating an adaptation degree score, generating personalized recommendation for the customer, preferentially recommending a product or service with the highest matching degree with the customer demand, and pushing related recommendation through multiple channels. And the satisfaction degree and the conversion rate of customers are improved. According to the invention, the resource utilization efficiency and the customer satisfaction are improved, so that more efficient enterprise operation and customer service are realized.

The invention relates to the technical field of unit fault diagnosis, in particular to a hydroelectric generating set fault diagnosis method and system based on digital twinning, and the method comprises the steps: dividing starting stages of a target hydroelectric generating set, analyzing starting transfer-in interference terms, building a filtering channel, evaluating the interference degree, combining real-time data anomaly analysis, and building an early warning model through a sensor algorithm; a digital twinborn model is established based on equipment topology, and a potential fault is predicted in combination with an early warning model, so that the technical problems that temporary normal fluctuation and a hydroelectric generating set fault are difficult to distinguish through an empirical threshold in a hydroelectric generating set starting process and the fault diagnosis analysis is limited are solved; effective separation and quantitative evaluation of various interference factors under different working conditions in the starting stage are realized, and the abnormality in the starting process is accurately captured by constructing a real-time data abnormality analysis module, so that the sensitivity of fault symptom detection is improved, a starting clamping stagnation early warning model is established, and the early warning efficiency is improved. And the fault diagnosis precision of the hydroelectric generating set is comprehensively improved.

The invention discloses a welding device and method for metal corrugated pipe production, and relates to the technical field of metal pipe production. A sliding rail is arranged at the top of the base, a collecting box is arranged on one side of the back of the sliding rail, two linkage plates are symmetrically arranged in the sliding rail, an L-shaped plate is arranged at the edge of the top of the base, an annular rail is arranged on the side, close to the axis of the base, of the L-shaped plate, and a welding assembly is arranged in the annular rail. Two deviation prevention devices are symmetrically arranged on the top of the base, and anti-falling devices are arranged on the inner sides of the deviation prevention devices. By means of the inclined plate and the round roller, it is guaranteed that a pipe entering the limiting ring is located in the center of the interior of the limiting ring all the time, the phenomenon of slight deviation in the pipe alignment process is avoided, and workers need to observe the deviated direction with the naked eyes so that manual adjustment can be conducted; and the alignment accuracy between the two pipes is effectively improved, and the welding preparation time is shortened.

The invention belongs to the technical field of artificial intelligence and natural language processing, and discloses a persuasion skill-based automatic generation large model jailbreak prompt method. According to the method, multiple persuasion skills are combined with multiple persuasion skills such as logic persuasion, emotional persuasion and authoritative persuasion, so that the attack success rate is increased; and dynamically adjusting the attack strategy: adjusting the attack strategy in real time according to the attack progress and target model feedback, continuously optimizing the attack prompt, and adopting different jailbreak methods for different target large models. And finally generating a prompt which enables various large models to break prisons with high probability based on various persuasion skills and feedback of the target large model, and providing a new data set for security testing of the large models.

The invention relates to the technical field of belt state detection, and particularly discloses a belt state detection method based on AI detection identification, and the method comprises the steps: belt deviation detection, deviation reason confirmation and feedback, belt wear detection, wear degree confirmation and feedback, and belt tear detection early warning. Whether the target belt deviates or not is judged according to the running video of the target belt, if the target belt deviates, the deviation reason of the target belt is determined, the abrasion degree of each area of the target belt is determined by constructing the three-dimensional information model of each area, meanwhile, tearing severity early warning is conducted by analyzing the tearing severity coefficient of the target belt, and therefore the tearing severity of the target belt is improved. The method improves the troubleshooting efficiency of belt deviation reasons, improves the accuracy of belt wear degree judgment, accurately evaluates the expansion direction and speed of cracks, provides effective data support for tear prevention and control, and helps to find potential tear risks in time, thereby reducing the possibility of belt tear accidents.

The invention provides a multimedia content AI detection method and device, a storage medium and electronic equipment. The method comprises the steps of performing preprocessing and feature extraction on input multi-modal data; extracting emotion and behavior risk features related to hidden illegal behaviors from the multi-modal feature representation; generating a novel violation template through zero sample learning, and matching the novel violation template with newly input multi-modal features to obtain a novel violation feature template and a matching score; performing semantic alignment on the novel violation feature template and the multi-modal features, and generating a preliminary violation risk score according to an alignment result and a matching score; and generating a structured report by integrating the grading result, the high-risk fragment index and the feature contribution degree. According to the innovative method and system, efficient detection and accurate recognition of the multi-modal violation content are achieved by combining emotion decoding, user behavior analysis and zero sample learning technologies, and therefore the defects in the prior art are overcome.

The invention relates to the technical field of computer data processing, in particular to a multi-model collaborative distillation and dynamic fine tuning model training method and system. The method comprises the steps that data collection is conducted, unlabeled data are input into a model to be made into labeled data, the labeled data and labeled data are integrated into a data set, and then the data set is divided into a training set, a verification set and a test set; inputting the training set into the model, guiding the model to generate extended data by cue words, and training a student model by using an extended training set; the student model is evaluated on the verification set, if the evaluation result shows that the evaluation index of the student model does not reach the expected value, error information is input into the model for analysis and data generation, and then the student model is adjusted until the evaluation index of the student model reaches the expected value; the method has the advantages that multiple large language models are used for collaborative distillation, training data are richer and more diversified, student models are more accurate, and the student models are circularly and dynamically fine-tuned, so that the models are higher in precision and adaptability.

The invention discloses a cultural heritage digital display platform, which is mainly applied to the fields of museums, cultural heritage protection institutions, education institutions and the like, and specifically comprises a central control module, and a user interaction module, a multimedia display module, a virtual reality rendering module, a database management module and an intelligent recommendation module which are connected with the central control module. The cultural heritage digital display platform has remarkable beneficial effects in the aspects of improving cultural heritage protection and inheritance efficiency, enhancing user experience and interactivity, realizing personalized recommendation and accurate service, promoting cultural education and propagation, optimizing data management and analysis capability and the like, and is beneficial to protection and inheritance of the cultural heritage; and the cultural literacy and aesthetic level of the public can be improved, and a positive contribution is made to the development of cultural heritage.

The invention discloses a task scheduling method and device, computer equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring rescue request information of all accident points; determining all schedulable rescue teams, and obtaining position information of each schedulable rescue team; calculating a rescue distance between each accident point and each dispatchable rescue team based on the rescue position information of each accident point and the position information of each dispatchable rescue team; constructing a rescue distance array based on the rescue distance between each accident point and each dispatchable rescue team; constructing an emergency level array and a resource demand array; and according to the rescue distance array, the emergency level array and the resource demand array, performing task scheduling on each schedulable rescue team by using a pre-configured greedy scheduling strategy. The invention further relates to the technical field of block chains, and the rescue request information is stored on a block chain network. The rescue efficiency and coverage can be comprehensively improved.

The present invention belongs to the technical field of construction risk assessment, and specifically discloses a subway construction safety risk intelligent assessment system, which includes a field data acquisition module, an equipment data import module, a construction data analysis module, a construction risk assessment module, a risk sorting and display module, a database, and an early warning execution and feedback terminal. The present invention solves the current deficiencies in data collection scope and depth by setting up key areas, and performing real-time construction site data acquisition and importing the accumulated construction time and real-time equipment monitoring data of key areas. At the same time, risk assessment is performed through three key aspects of personnel, equipment and structure, forming an independent and complete risk indicator analysis system, and by constructing risk indication maps in each key area, it is convenient for construction management personnel to conduct comprehensive risk monitoring and management, and improves the effectiveness of risk control and feedback links.

The invention provides a software system dependency defect tracking method and system based on a graph database, and relates to the technical field of software, and the method comprises the steps: parsing a source code into an abstract syntax tree, and extracting a dependency; constructing a dependency graph data model; establishing a defect mode feature library; performing pattern matching in the graph data model to identify potential defects; calculating a defect influence range and severity; and generating a defect tracking report, including defect priority ranking, complete call chain construction, defect detailed information recording and optimization suggestion generation. According to the method, the dependency relationship defect in a software system can be effectively identified and tracked, and the software quality is improved.

The utility model discloses a banknote separating device and financial equipment, and belongs to the technical field of financial equipment. The banknote separating device comprises a banknote supporting plate; the mounting frame and the banknote supporting plate are distributed in a spaced mode, and a banknote inlet channel is defined; the impeller is installed on the installation frame, the distance between the axis of the impeller and the end portion of a blade of the impeller is larger than the vertical distance between the axis of the impeller and the banknote supporting plate, and the blade is a flexible material piece. And the driving mechanism is in dynamic coupling connection with the impeller and is used for driving the impeller to rotate. According to the banknote sorting device provided by the invention, the impeller is utilized to push the financial medium to move towards the direction of entering the banknote inlet channel, and the contact range with the financial medium can be enlarged, so that the financial medium close to the inlet of the banknote inlet channel can be conveniently pushed towards the outlet of the banknote inlet channel, and the banknote sorting effect and the banknote sorting efficiency of the banknote sorting device are improved.

The invention belongs to the technical field of construction risk assessment, and particularly discloses and provides a subway construction safety risk intelligent assessment system, which comprises a field data acquisition module, an equipment data import module, a construction data analysis module, a construction risk assessment module, a risk arrangement display module, a database and an early warning execution and feedback terminal. According to the method, the key area is set, real-time construction site data acquisition is carried out, and accumulated construction duration and real-time equipment monitoring data of the key area are imported, so that the problem of deficiency in the aspects of data acquisition range and depth at present is solved, and risk assessment is carried out through three key aspects of personnel, equipment and structure, so that the risk assessment accuracy is improved. An independent and complete risk index analysis system is formed, and the risk indication diagrams in the key areas are constructed, so that the construction management personnel can conveniently carry out comprehensive risk monitoring and management, and the effectiveness of risk management and control and feedback links is improved.

The invention belongs to the technical field of intelligent questioning and answering, and particularly discloses a business questioning and answering method and system based on a large language model and a scene label, corresponding knowledge base question sets are associated through business question expansion matching, and then answers are called from a knowledge base to determine corresponding data blocks and scene labels. Then, through scene label matching and scene consistency evaluation, a database with high scene consistency is selected to be combined with user questions to be input into a large language model to achieve intelligent question and answer. The business question and answer requirements of the user can be effectively met, and the user does not need to directly query professional knowledge; according to the method, questions proposed by users can be analyzed more comprehensively, questioning intentions and question scene information are fully covered, and question and answer feedback results are more targeted and have higher relevance with business questions of the users; and the question and answer result is simpler and more effective.

The present invention discloses a transcranial magnetic stimulator, which belongs to the technical field of medical devices. It includes: a circular hole is provided on a substrate; a first rotating ring is rotatably installed in the circular hole of the substrate; a circular hole arranged eccentrically relative to the axis of the first rotating ring is provided on the first rotating ring; a first driving mechanism is installed on the substrate, which is used to control the first rotating ring to rotate around the circular hole of the substrate; a second rotating ring is rotatably installed in the circular hole of the first rotating ring; a circular hole arranged eccentrically relative to the axis of the second rotating ring is provided on the second rotating ring; a second driving mechanism is installed on the first rotating ring, which is used to control the second rotating ring to rotate around the circular hole of the first rotating ring; a shaft cylinder is rotatably installed in the circular hole of the second rotating ring; a magnetic head is installed at the lower end of the shaft cylinder. The present invention realizes the flexible movement of the magnetic head on a two-dimensional plane through a two-stage rotating ring, which greatly increases the flexibility and accuracy during treatment, helps to achieve more accurate and personalized treatment, and improves the treatment effect.

The invention discloses an agricultural disease and pest dialect speech intelligent recognition method based on a deep learning technology, and the method comprises the steps: combining a Mel-frequency cepstral coefficient with a convolutional neural network, and dividing a whole process into two stages, the pre-training stage comprises the steps of recording and storing a long voice file, extracting the characteristics of voice signals in the file and training a neural network model, and the retraining stage comprises the steps of recording and storing a short voice file, extracting the characteristics of the voice signals in the file, verifying a test result, adjusting model parameters and storing an optimal model; the dialect voice information reported by monitoring personnel in an agricultural field is intelligently recognized by considering accent of people in different regions, so that the recognition accuracy of complex dialect voice signals is effectively improved, and the recognition method is particularly suitable for understanding related terms of agricultural diseases and insect pests.

The invention provides a slope cropland fertilizing device and method, belongs to the technical field of fertilizing, and aims to solve the problem that a fertilizing device special for slope cropland is lacked in the prior art. Comprising a vehicle body, a chemical fertilizer box, a fertilizer distributor, a mixer, a fertilization angle adjusting assembly and a fertilization gun, a power device is arranged on the vehicle body, the vehicle body is driven by the power device to advance, and a handrail is fixedly arranged on the back of the vehicle body; the chemical fertilizer boxes are symmetrically and fixedly arranged on the two sides of the vehicle body, through holes are formed in the bottoms of the chemical fertilizer boxes, the through holes in the bottoms of the chemical fertilizer boxes are communicated with the fertilizer distributors, and the fertilizer distributors are used for quantitatively measuring chemical fertilizers; the bottom of the fertilizer distributor is communicated with the mixer; the fertilization angle adjusting assembly is arranged at the bottom of the vehicle body; the fertilization angle adjusting assembly is used for adjusting the fertilization angle; the fertilizing gun is mounted on the fertilizing angle adjusting assembly.

The invention relates to the technical field of hydraulic brake joint detection, and discloses a hydraulic brake joint thread automatic detection device, which comprises a feeding bottom plate arranged on an annular conveying line, a first detection station and a second detection station, a side plate is fixed on the feeding bottom plate, the side plate is rotatably connected with a rotating shaft, and the first detection station and the second detection station are arranged on the feeding bottom plate. A rotating shaft is arranged on the side plate, an overturning seat is fixed at the end part of the rotating shaft, a first rotating groove is formed in the side plate, a gear is rotationally arranged in the first rotating groove, a first toothed plate is meshed on the gear, and a first slide way for the first toothed plate to slide is formed in the side plate; according to the technical scheme, air can be blown into the joint workpiece before the two ends of the joint workpiece are detected, foreign matter in the joint workpiece can be conveniently removed, the detection accuracy can be improved, meanwhile, the overturning base can be overturned, the two ends of the joint workpiece can be conveniently detected, and the detection efficiency is improved.

The invention discloses an intelligent touch system and method based on user characteristics, and belongs to the technical field of message pushing. According to the system, user portraits are generated through user behavior data, the user portraits comprise tags such as reach channels, reach time and reach content, and preference characteristics such as channel preference, time preference, algorithm preference and copywriting preference are generated based on click behaviors of users. And the system calculates the weight of each reaching channel, time and content packet through an element preferential module, and selects an optimal combination to generate a message sending scheme. And if the message is not successfully sent, the system automatically selects to generate a new sending scheme again and sends the new sending scheme again. According to the method, accurate and personalized message pushing can be realized according to historical behaviors and preferences of the user, the reaching rate of the message and the response rate of the user are remarkably improved, and the method is suitable for scenes needing large-scale personalized pushing such as e-commerce platforms and social media.

The invention discloses a tourism advertisement marketing system based on big data, and relates to the field of advertisement putting, and the system comprises the following components: an intelligent putting and social promotion module which accurately puts an advertisement to a target audience according to an analysis result of a user relation network, and simultaneously puts the advertisement to the target audience through sharing, forwarding and interaction functions of a social media platform; according to the system, a user can be guided to participate in propagation and promotion of advertisements, the system can also adjust a putting strategy in real time according to feedback and interaction data of the user, the propagation effect and conversion rate of the advertisements are improved, and through a reinforcement learning algorithm adopted by the intelligent putting and social promotion module, the characteristics of social media are fully utilized; through sharing, forwarding and interaction functions, the advertisement can be quickly spread in the user relation network, so that the spreading range of the advertisement can be expanded, the number of potential customers can be increased, the credibility and persuasion of the advertisement can be improved, and the advertisement can be more easily trusted and recognized by other users.

The present invention relates to an aerated biological filter with good purification effect, comprising: a tank body, a filter material layer arranged in the tank body, a water outlet pipe arranged at the top of the tank body, a water inlet pipe arranged at the bottom of the tank body, and an aeration mechanism arranged below the filter material layer, wherein the aeration mechanism comprises: a fixing frame fixed in the tank body, an aeration bracket rotatably arranged on the fixing frame, an air inlet pipe connected to the bottom of the aeration bracket, and a rotary drive motor. The aeration mechanism of the present invention is provided with an aeration bracket with an air inlet cavity arranged inside, wherein a plurality of air outlet rods connected to the air inlet cavity are arranged around the top of the aeration bracket, and a plurality of air outlet holes are staggeredly distributed on each air outlet rod, and the aeration bracket is driven to rotate by a rotary drive motor, so that the floating bubbles formed by the plurality of air outlet rods continuously change the floating path, thereby improving the purification effect of the aerated biological filter on sewage.

The invention relates to the technical field of pet food processing, and discloses surrounding hot air slow drying equipment for pet food processing and a three-section type baking method.The surrounding hot air slow drying equipment for pet food processing comprises a drying box, a hot air drying unit is installed at the bottom in the drying box, and a circulating conveying unit is installed in the drying box; a plurality of drying frames are evenly distributed on the circulating conveying unit, material overturning mechanisms are arranged in the drying frames, positioning assemblies are installed in the drying box, and the material overturning mechanisms and the positioning assemblies are matched to be used for overturning raw materials in the drying frames. Food particles on the drying frame are turned over by moving a turning plate in the drying frame, meanwhile, an arc-shaped plate and a second fixing plate are matched to enable the food particles to be scattered in an S shape, damage to the food particles is avoided, meanwhile, the contact face of the food particles and hot air is replaced, and the drying effect is improved while the pet food yield is increased; through the three-stage baking process, the gelatinization degree of the baked grain starch of the pet food can be increased to 90% or above, and digestion and absorption are facilitated.

The invention discloses an evaporation test pool for monitoring water surface evaporation data, and belongs to the field of meteorological detection.The evaporation test pool comprises an evaporation mechanism and a floating mechanism, the evaporation mechanism is located on the lower portion of the inner side of the floating mechanism, the evaporation mechanism comprises a pool body, a plurality of partition plates are fixed in the pool body, the partition plates are fixedly connected, and the floating mechanism is arranged in the pool body. Through the connection of the screw rod and the screw cylinder, the distance and height difference between the pool body and the floating mechanism can be adjusted, so that the depth of the pool body in water is controlled. When the electromagnetic valve is opened to enable the tank body to be communicated with water, the water level in the tank body can be accurately controlled; through a series of mechanical transmission structures such as a wave plate, a connecting rod, a toothed plate, a ratchet wheel and a pawl, when the water surface fluctuates, the test pool can automatically sense and convert wave energy into mechanical energy, the floating mechanism is driven to expand or contract, the contact area with the water surface is increased, the wind wave resistance is improved, and meanwhile the stability of the equipment is guaranteed.

The invention provides an intelligent monitoring system and method for well drilling overflow well leakage, and relates to the field of intelligent monitoring, and the method comprises the steps: firstly, collecting the mass flow at an inlet and an outlet of a well drilling circulation system, transmitting the mass flow to a data processing cloud, and then carrying out the feature extraction of the collected data through employing a data analysis and processing algorithm based on artificial intelligence, and the overflow condition and the well leakage condition of the well drilling are monitored by using the time sequence collaborative change between the mass flow time sequence data at the inlet and the mass flow time sequence data at the outlet. Thus, by comprehensively analyzing the time sequence characteristics of the inlet and outlet mass flow, the dynamic relevance of flow change in the drilling process can be comprehensively captured, the monitoring coverage is improved, and therefore the intelligent degree of drilling overflow well leakage monitoring is improved, and the safety and efficiency of drilling operation are improved.

The invention discloses an intelligent cable inspection system and inspection method, and belongs to the technical field of cable inspection, and the system comprises an Internet of Things module, and an automatic inspection module, a data collection module, a data monitoring module, a continuous monitoring module, a data background storage module and a visualization module which are wirelessly connected through the Internet of Things module. The automatic inspection module comprises a robot carrying an infrared temperature sensor and an unmanned aerial vehicle, the data acquisition module comprises a digital voltage sensor, the continuous monitoring module comprises a humidity sensor, a Hall effect sensor and a direct current leakage current sensor, and the data background storage module comprises a data recording module, a data storage module and a data analysis module. The cable inspection efficiency and accuracy are improved through the processes of automatic inspection, data acquisition, data monitoring, data storage and visualization.

The invention discloses a smart park digital management system based on a digital twinning technology, which comprises a function management system, a digital twinning system and a data center, and is characterized in that the function management system is used for comprehensive function management and information acquisition of a park, and the digital twinning system is in communication association with the function management system; related digital twinborn models are respectively constructed according to security and protection information, fire alarm information, equipment information, energy consumption information, personnel and vehicle information acquisition, internet-of-things interaction and park data information transmitted by the function management system; and the data center adopts a distributed storage strategy and is used for providing basic information, equipment information, data information and model information of operation of the function management system and the digital twin system. According to the invention, on the basis of modern information technologies such as the Internet of Things, cloud computing, digital twinning and the like, a unified organization management coordination architecture and a business management system are constructed in the aspect of informatization construction of the park, and innovative management and operation services are provided for park managers and enterprises.

The invention relates to the technical field of digital human systems, and discloses an enterprise propaganda virtual digital human system based on a block chain and deep learning, and the system ensures the safety and non-tampering of enterprise propaganda data and user interaction data through a block chain storage unit; the intelligent level of the virtual digital human is improved by utilizing a deep learning model construction unit, and the functions of language understanding, dialogue generation, image rendering and the like are realized; the enterprise information extraction unit and the user request processing unit are respectively responsible for data updating and request analysis; the dialogue management unit and the image rendering unit generate proper replies and dynamic images according to user requirements and scenes; the block chain verification unit guarantees the authenticity of the data; the feedback optimization unit realizes continuous optimization of the model; and the multi-channel publishing and managing unit ensures multi-platform synchronization and consistency of contents. According to the system, the personalized, interactive and efficient levels of enterprise propaganda are greatly improved.

The invention relates to the technical field of urban pipe network detection, in particular to a pipe diameter self-adaptive pipeline interior detection robot which comprises a main machine box, a sliding rod, a rotary drum, a transmission assembly A, a driving assembly, an annular mounting shell, a transmission assembly B and an oil pumping mechanism. The two ends of the mainframe box are each connected with a guide head, an annular oil cavity is formed in each guide head, and a guide pipe is arranged on each guide head. And the other end of the sliding rod is connected with the roller. The rotary drums are inserted into the guide pipes on the corresponding sides respectively, and spline shafts are arranged in the rotary drums and inserted into the sliding rods. The transmission assembly A is in transmission connection with the spline shaft and the rotating shaft. The annular mounting shell sleeve is connected with the mainframe box, and an X-ray flaw detector is arranged on the annular mounting shell. The transmission assembly B is in transmission connection with the annular mounting shell and the driving assembly. The oil pumping mechanism is communicated with the annular oil cavities in the guide heads on the two sides. The device is high in detection efficiency and low in energy consumption, and synchronous adjustment of the distance between the rollers on the two guide heads is achieved through one oil pumping mechanism.

The utility model discloses a high-temperature-resistant glass processing edging device which comprises a base, a shell is arranged above the base, the shell and the base are welded and fixed, a workbench is arranged on one side of the upper end face of the base, and a supporting seat is fixedly arranged on one side in the shell. The multi-shaft adjusting mechanism is arranged on one side of the supporting seat, the multi-shaft adjusting mechanism comprises a lifting seat, the lifting seat is slidably connected with the supporting seat, a first adjusting seat is installed on one side of the lifting seat, a second adjusting seat is installed on one side of the first adjusting seat, and the second adjusting seat is rotatably connected with the first adjusting seat; a driving seat is mounted below one side of the second adjusting seat, and the driving seat is rotationally connected with the second adjusting seat; and the rotating base is arranged in the middle of the upper end face of the workbench, a vacuum suction cup is fixedly installed above the rotating base, and the problem that grinding dead angles are likely to occur in the machining process of kitchen ware glass is solved.

The present invention discloses an outdoor advertising recommendation method based on large-scale user portraits, and belongs to the technical field of computational advertising. The advertising scheduling module in the outdoor advertising recommendation method of the present invention is respectively connected to the user end, the advertising management module, and the real-time and accurate crowd analysis module. The real-time and accurate crowd analysis module performs parameter interaction with the advertising retrieval module, the user portrait module, and the user behavior module respectively. The user data mining module is respectively connected to the advertising management module, the user behavior module, and the user portrait module, and the advertising management module is connected to the advertising retrieval module. The outdoor advertising recommendation method of the present invention completes the generation of a personalized advertising list based on the user portrait query of the user's interests and historical behavior records when the user logs into the platform, and finally pushes the advertisement to the outdoor advertising terminal. The present invention has good autonomous learning ability, can effectively improve the reach rate of outdoor smart advertising, and is suitable for outdoor advertising recommendation under the background of the Internet of Things and big data.