517results about How to "Adjust in time" patented technology

The invention discloses an industrial product quality detection method and system based on machine vision, and relates to the technical field of quality detection.During operation of the system, through an industrial camera and an optimized light source layout, high-resolution image acquisition is performed on electronic components in real time by adopting multi-angle shooting, a series of optimization processing is performed on the acquired images, and the quality of the electronic components is detected. The method comprises the following steps: extracting shape, texture and color features according to geometric characteristics of an electronic component to form key parameters for judging defects, identifying potential defects of the electronic component by using a deep learning model algorithm, calculating to obtain a quality evaluation index Q, and calculating the quality evaluation index Q according to calculation results of feature extraction and defect detection. And comprehensively evaluating the quality of each component, comparing the quality evaluation index Q with a first preset threshold value M and a second preset threshold value N to obtain a quality evaluation grade, feeding unqualified detection information back to a production line, and removing unqualified components.

The invention discloses a method for realizing high-efficiency low-consumption micro-aerobic hydrolytic acidification of petrochemical wastewater by regulating and controlling the aeration rate in combination with a space-time diagram neural network and reinforced learning, and relates to a method for realizing high-efficiency low-consumption micro-aerobic hydrolytic acidification of petrochemical wastewater by regulating and controlling the aeration rate. The method comprises the following steps of: 1, collecting and cleaning data; 2, constructing a graph data set; 3, training the ST-GNN model; 4, constructing a deep Q reinforcement learning model; 5, the trained ST-GNN model is embedded into an automatic control system of a sewage treatment plant, and the model outputs the optimal aeration rate of the current time step and predicts the micro-aerobic hydrolytic acidification effect through interaction of the current working condition and historical data; and adjusting operation parameters of aeration equipment according to the output of the model, and carrying out hydrolytic acidification treatment. The method is high in regulation and control efficiency and low in operation error, and can be applied to the field of digital and intelligent management of sewage treatment plants.

The invention provides a front end assembly, an endoscope and an endoscope hydraulic control system, and belongs to the technical field of endoscopes. The front end assembly comprises a front end base, a camera module and a pressure measuring mechanism, the camera module and the pressure measuring mechanism are both installed on the front end base, the pressure measuring mechanism comprises a piezochromic layer, the piezochromic layer is located at the circumferential position of the camera module, the piezochromic layer is made of a piezochromic material, and at least part of the piezochromic layer is located in an image acquisition area of the camera module. The front end seat is provided with the pressure measuring mechanism, and the pressure measuring mechanism is used for detecting the pressure around the front end assembly, so that the accuracy of pressure adjustment in a body cavity is improved, the hysteresis of pressure adjustment in the body cavity is reduced, the amplitude of pressure fluctuation in the body cavity is reduced, and the surgical risk is reduced; the pressure information in the body cavity is judged by observing the color information of the piezochromic layer in the image information collected by the camera module, an operator can adjust the pressure in the body cavity in time, and the overall operation is simple and convenient.

The invention provides an air conditioner personalized intelligent control method, system and device and a storage medium, and relates to the technical field of air conditioner control, the method comprises the following steps: combining first people flow data and historical people flow data of a target area within a first preset duration to predict second people flow data within a target duration; determining the reference temperature of the target area within the target duration in combination with the outdoor temperature and the area characteristics of the target area; in combination with the reference temperature and the second visitor flow rate data, temperature adjusting parameters of the target area within the target duration are determined, and an initial operation scheme of an air conditioner of the target area is formulated according to the temperature adjusting parameters; and according to the environment monitoring parameters of the target area, determining an environment load coefficient, and according to the environment load coefficient, generating a target operation scheme. The method has the technical effects that the environment change is predicted in advance, the air conditioner operation parameters are adjusted in time, and it is guaranteed that the indoor environment can reach the comfortable state in time.

The invention provides a manufacturing-oriented test process digital twinning and dynamic optimization management method, which relates to the technical field of testing, and comprises the following steps of: acquiring real-time test data through a multi-source data acquisition device, and establishing a test process digital model; constructing a test process digital twin system; establishing a test process prediction model by using a deep learning algorithm; comparing and analyzing the real-time data and the prediction result, and identifying abnormity and optimization opportunities; generating and implementing a test process optimization scheme; and closed-loop optimization management is realized through feedback control. According to the method, the efficiency and quality of the test process are improved.

The invention relates to a wheel type double-arm humanoid robot motion track planning and obstacle avoidance method. The method comprises the steps that a robot system is started for self-inspection; the first sensor obtains environment information and constructs a real-time three-dimensional point cloud obstacle information graph; the second sensor obtains chassis state information and mechanical arm state information; a target task is set, and a moving path of a chassis and a moving path of a mechanical arm are planned based on the state information of the target task; and the system controls the chassis to move according to the planned moving path through the cooperative control model, and the mechanical arm moves according to the planned moving track. Through construction and real-time updating of the real-time three-dimensional point cloud obstacle information graph, the robot can effectively recognize and avoid obstacles, the risk of collision with the surrounding environment is reduced, meanwhile, the robot can adapt to the dynamically changing environment, path planning is rapidly updated through the D * Lite algorithm, and the path planning efficiency is improved. And the path can be adjusted in time when the obstacle moves or a new obstacle appears.

The invention discloses an intelligent rehabilitation treatment management system and method based on big data, and relates to the technical field of intelligent management. During operation of the system, rehabilitation-related data of an orthopedic patient is collected, real-time data is obtained through intelligent wearable equipment and a sensor, the pain and fatigue state of the patient is monitored, preprocessing is carried out, and the patient is subjected to rehabilitation treatment; the method comprises the following steps: analyzing collected data, monitoring the rehabilitation progress of a patient in real time, calculating to obtain a comprehensive rehabilitation progress evaluation index Etotal, comparing the comprehensive rehabilitation progress evaluation index Etotal with a preset state threshold value M and a second preset state threshold value N to generate a personalized rehabilitation scheme, and remotely monitoring rehabilitation data of the patient by a doctor so as to obtain a personalized rehabilitation scheme. According to the method, the rehabilitation plan and the training progress of a patient can be checked through a mobile phone APP, the rehabilitation plan, the training progress and real-time feedback are fed back, intelligent reminding is obtained, and through rehabilitation video or voice guidance, the patient can share experience, exchange rehabilitation experience and obtain psychological support and encouragement.

The invention discloses an anti-carsickness semi-active seat band-stop filtering control method, which comprises the following steps of: firstly, establishing a two-degree-of-freedom model of a vehicle seat suspension system, and designing a seat suspension active control system to meet the following constraint conditions: the dynamic stroke of a seat suspension does not exceed the maximum allowable stroke; and the control force is smaller than the maximum output control force of the actuator. In order to effectively reduce the carsickness symptom caused by seat vibration, the invention also establishes a target function for inhibiting the vibration of the seat in a human body sensitive frequency range, designs a frequency weighted sliding mode control algorithm, accurately adjusts the damping coefficient of the magnetorheological damper through a control law, inhibits the low-frequency vibration in the frequency range of 0.1-0.5 Hz, and improves the safety of the seat. The riding comfort is improved, and the motion sickness is reduced.

The invention discloses a ship dangerous goods transportation safety control method and system, and relates to the field of marine transportation, and the method comprises the steps: obtaining dangerous goods parameter data of a dangerous goods monitoring sensor, the dangerous goods parameter data comprising gas concentration, temperature and smoke concentration; under the condition that the sensor parameter data exceeds a preset parameter threshold value, acquiring motion attitude data and inclination angle data of the current ship and pressure distribution data of the cargo on the bottom and the side surface of the cargo hold; calculating the center-of-gravity offset of the current ship based on the motion attitude data, the inclination angle data and the pressure distribution data; calculating a fixed adjustment parameter of the cargo fixing device based on the gravity center offset; and sending a device adjustment instruction containing the fixed adjustment parameter to the target client. By implementing the method, the efficiency of adjusting the cargo fixing device can be improved.

The invention discloses a water quality self-adaptive regulation and control system and method based on multi-point detection, and relates to the technical field of water quality self-adaptive regulation and control. A sensor network collects water quality data in each sub-region of a target water area in real time, and abnormal regions are screened out after water quality abnormal states in the sub-regions are determined according to water quality abnormal scores; performing inspection on each abnormal area by using an inspection unmanned aerial vehicle, and summarizing to generate a water area inspection data set after obtaining corresponding inspection data; outputting an intervention strategy for the target water area by the water quality intervention knowledge graph, executing the intervention strategy, automatically controlling the operation of the water treatment facility by using the automatic control model, collecting corresponding intervention feedback data, and adjusting and optimizing the intervention strategy according to the intervention feedback data; and a water quality monitoring report is sent to a data receiving end at the time node meeting the constraint condition. When the water quality in the target water area is abnormal, automatic and adaptive adjustment control is realized, and the water quality in the target water area is prevented from deteriorating.

The invention belongs to the technical field of network self-adaption, and particularly relates to a weak network self-adaption algorithm system based on dynamic network monitoring, which comprises the following steps of: deploying a monitoring agent on a network node; extracting network bandwidth, delay and packet loss rate from the network state report, preprocessing the extracted data, and extracting feature vectors from the preprocessed data; adjusting the size of a buffer area for data transmission according to the network bandwidth and delay; and controlling the sending and receiving of data according to the adjusted algorithm and parameter, monitoring the data transmission state in real time in the data transmission process, feeding back the data transmission state information, and adjusting the algorithm and parameter in time according to the feedback information. The problem of how to improve the efficiency and reliability of data transmission under the conditions of complex and changeable network environment, low bandwidth, high delay, serious packet loss and the like in the prior art is solved.

The invention relates to an inspection robot global path planning method based on a three-dimensional grid map. The method comprises the following steps: constructing a three-dimensional point cloud map; constructing a three-dimensional occupation grid map; calculating the height difference between each grid and an adjacent grid of the three-dimensional occupied grid map, and marking the current grid and continuous grids in the vertical direction, namely the z direction, as grids with the same attribute; giving current position and target point position information of the robot, and converting the information into serial numbers to which grids belong; and traversing surrounding grids from a starting point by using an A * algorithm until a target point position is searched, and forming a global path from the starting point to the target point. According to the invention, a traditional two-dimensional grid map is replaced by a three-dimensional grid map, so that the obstacle avoidance capability of the ground mobile robot in a complex terrain is remarkably improved, the flexibility of the robot is improved, the safety of the robot in an actual application scene is enhanced, and especially in a dynamically changing environment, the obstacle avoidance capability of the robot is greatly improved. The path can be adjusted in time to avoid collision.

The invention relates to the technical field of wastewater defluorination control, in particular to a wastewater defluorination control system suitable for a self-crystallization fluidized bed, which comprises a wastewater defluorination platform, an information acquisition unit, a defluorination state unit, a dosing supervision unit, a treatment matching unit, a tracking supervision unit and a management response unit, according to the method, analysis is carried out from two points of the target equipment and dosing, so that the defluorination efficiency of the target equipment is ensured, defluorination demand matching supervision analysis is carried out on the processing data in an information feedback mode, so that a defluorination scheme is reasonably matched according to the data processing condition, resource waste is avoided, the operation cost is reduced, and the method is suitable for popularization and application. Meanwhile, dosing control supervision feedback analysis is deeply carried out, on one hand, stirring interference of dosing is adjusted so as to ensure full contact reaction of fluorine ions in the wastewater, and on the other hand, the dosage is adjusted so as to ensure the fluorine removal efficiency of the wastewater and the rationality of automatic dosing control of chemicals.

The invention provides laser cutting feeding equipment, and relates to the technical field of laser cutting equipment, the laser cutting feeding equipment comprises a feeding connecting piece, and the feeding connecting piece is provided with a row of lifting supporting pieces; two magnetic control pieces are respectively mounted at the bottoms of the row of lifting support pieces; the two rows of magnetic control pieces are respectively arranged on the feeding connecting piece; the row of lifting supporting pieces are used for supporting to-be-machined plates correspondingly. A path control piece is mounted on the feeding connecting piece; the path control piece is used for controlling the row of lifting supporting pieces to descend to prevent shielding, can support a to-be-machined plate in a rolling mode, prevents abrasion, can automatically descend, prevents adhesion of cutting melt and avoids deformation caused by laser heating, and solves the problems that a feeding sawtooth plate in existing laser cutting feeding equipment is prone to scratching the plate, and the production efficiency is high. And a large number of impurities are easily adhered to the sawtooth plate, so that the sawtooth plate is inconvenient to automatically return to a cutting position according to a cutting path.

The invention discloses a mattress roadway stacking machine for bamboo mattress production and processing, and relates to the technical field of automatic storage and logistics, the mattress roadway stacking machine comprises a frame, a base plate, a conveying belt, a mattress, a stacking frame and a fixing frame, a stacking device is arranged in the stacking frame, a conveying device is arranged above the conveying belt, and the fixing frame is arranged above the conveying belt. An auxiliary device is arranged above the fixing frame, a base plate is pushed to move along a sliding rail rod through a servo motor, a conveying belt and a push plate, mattresses can be more easily aligned and stacked in the follow-up stacking process through the base plate, the base plate is moved upwards through an air cylinder and a top plate, the base plate is placed at the top of the placing frame through the placing frame and a check rod, and therefore the mattresses can be conveniently stacked. Stacking of the base plate is completed at the moment, stacking of the base plate is gradually conducted upwards, stacking of the mattresses can be more stable, the collapse risk is reduced, storage and management of the mattresses can be more orderly through the stacking mode, and the mattresses are convenient to check and take for use.

The invention discloses an energy efficiency optimization method and system for an irrigation area, a storage medium and electronic equipment, and relates to the field of intelligent irrigation. The method comprises the steps that crop information and soil information of all sub-areas in an irrigation area are obtained, and the irrigation area is communicated with branch pipelines of all the sub-areas through a main pipeline; for any sub-region, according to the crop information and the soil information, determining the irrigation amount and the irrigation priority required by crops in the sub-region; according to the irrigation amount and the irrigation priority corresponding to each sub-region, a first irrigation plan of a connected region is generated, the connected region is composed of at least two sub-regions, and the first irrigation plan comprises an irrigation mode for irrigating crops in each sub-region in the connected region in different time periods. By implementing the technical scheme provided by the invention, the overall energy efficiency optimization of the irrigation system can be realized.

The invention provides a hotel guest room control system and method based on big data analysis, and relates to the field of intelligent control. Temperature time sequence data, humidity time sequence data, carbon dioxide concentration time sequence data and PM2.5 concentration time sequence data of hotel guest rooms are obtained through sensors deployed in the hotel guest rooms; and the time sequence data of the air quality parameter indexes are analyzed to know the air quality condition in the hotel guest room, and the operation of the fresh air system is adaptively controlled based on the air quality condition in the hotel guest room. In this way, timely and accurate adjustment can be made according to the actual air quality condition in the guest room, and therefore the flexibility and the response speed of the fresh air system are improved.

The invention belongs to the technical field of enterprise digitization, and particularly relates to a rapid tracking and positioning method for enterprise digitization development maturity. Comprising the following steps: S1, establishing a digital development maturity evaluation framework; s2, index system screening is carried out based on a fuzzy Delphi method; s3, establishing an improved K-means model based on density and weight; s4, carrying out enterprise digital development grade division; s5, establishing an enterprise digital transformation mature stage correction model; according to the method, a digital development maturity evaluation framework is constructed, index system screening based on a fuzzy Delphi method is utilized, an improved K-means model based on density and weight is combined, enterprise digital development grade division and an enterprise digital transformation maturity stage correction model are combined, enterprises are helped to quickly know the current situation of own digital development, and the evaluation accuracy of the digital development maturity is improved. Problems can be found in time and targeted measures can be taken so that a problem that rapid and accurate tracking and positioning of enterprise digital development maturity are lacked in the prior art can be solved.

The invention relates to the technical field of gas supply, and discloses a steel ladle bottom blowing slow-release gas supply structure and method. The control module is arranged on the steel ladle main body and is used for being responsible for automatic control including sensor data monitoring and valve opening adjusting; the sensor module is arranged on the steel ladle body and used for monitoring temperature, pressure and gas flow parameters in real time and feeding data back to the control module; the gas supply module is used for conveying through high-pressure air or other inert gases and providing a stable gas source for the bottom blowing device of the steel ladle main body; and the slow release device is arranged at the bottom of the steel ladle main body and is used for slowly releasing the supplied gas so as to ensure the uniform distribution of the supplied gas. According to the steel ladle bottom blowing slow-release gas supply structure and method, the efficiency and safety in the smelting process are improved, the product quality is improved, resource saving and intelligent management are achieved, and important support is provided for development of the modern steel and iron manufacturing industry.

The invention relates to a variable-amplitude oil cylinder with a buffering function, and belongs to the technical field of hydraulic oil cylinders, the variable-amplitude oil cylinder comprises a cylinder body, a piston rod and a cylinder cover, a hollow ring is slidably arranged between the cylinder body and the piston rod, a partition plate is arranged in the hollow ring, and the interior of the hollow ring is divided into a liquid storage cavity a and a liquid storage cavity b by the partition plate; a displacement sensor used for monitoring the displacement distance of the piston rod in real time is arranged at the tail end of the cylinder body, a piston is arranged at the tail end of the piston rod, the front end of the piston is connected with a plurality of connecting rods, the front ends of the connecting rods are connected with annular push plates arranged in the liquid storage cavity a in a sliding mode, and when the piston rod stretches out forwards, the annular push plates push liquid in the liquid storage cavity a into the liquid storage cavity b. Reverse thrust of liquid abuts against thrust of hydraulic oil pushing the piston rod, so that the piston rod is buffered, and the piston rod can extend out gently in the starting stage. The problem that when the oil cylinder is started, the piston rod rapidly stretches out to generate shaking and instant impulsive force, and consequently precision mechanical equipment borne by the piston rod is potentially damaged is solved.

The invention discloses a roadway stacking machine for stacking packaging boxes, and relates to the technical field of packaging box storage, the roadway stacking machine comprises a connecting plate, an electric sliding table I is fixedly mounted on the upper surface of the connecting plate, a storage iron frame is fixedly mounted on the upper surface of the connecting plate, and a sliding block I is mounted at the sliding end of the electric sliding table I; a second electric sliding table is fixedly connected to the upper surface of the first sliding block, a second sliding block is installed at the sliding end of the second electric sliding table, and a correcting assembly is arranged on the bearing platform, so that the effect of ensuring that the position of the packaging box can be adjusted in time when the packaging box is placed on the bearing platform is achieved; the packaging box is accurately located in the center of the bearing platform, in the operation process of the roadway stacking machine, the packaging box cannot deviate due to improper placement position, meanwhile, the situation that the packaging box cannot be placed on the storage iron frame due to inclination is avoided, and therefore the risk that the packaging box falls off is reduced, and the service life of the packaging box is prolonged. And the economic loss caused by falling of the packaging box is reduced.

The invention relates to the technical field of medical image analysis, in particular to a training method and system for a cataract recognition model and a storage medium. The method comprises the following steps: acquiring cataract eye image samples and non-cataract eye image samples in different stages; lesion region segmentation and image synchronous simulation enhancement are carried out on the cataract eye image samples and the non-cataract eye image samples in different stages to obtain cataract lesion enhanced image samples and non-cataract enhanced image samples; extracting local detail features of the lesion and extracting structural features of a mask pair, and constructing a multi-task convolutional neural network learning framework to train a cataract recognition model and optimize focus weighted loss; and obtaining a cataract related key feature set and carrying out online incremental learning updating to generate a cataract recognition online incremental updating model. According to the invention, the accuracy and robustness of cataract recognition model training can be improved.

The present application relates to the technical field of hot-rolled strip thickness, and discloses a monitoring method and device for a hot-rolled thickness control model, the method comprising: obtaining a self-learning coefficient calculated and generated by a self-learning module in the hot-rolled thickness control model during multiple hot-rolling processes to obtain a first monitoring parameter array; judging whether there are J1 consecutive values ​​in the first monitoring parameter array whose absolute values ​​all exceed a set limit value, wherein J1 is a natural number greater than 2; if there are J1 consecutive values ​​in the first monitoring parameter array whose absolute values ​​all exceed a set limit value, then outputting alarm information about the hot-rolled thickness control model. The technical solution provided by the present application monitors whether there is a problem with the hot-rolled thickness control model through the self-learning coefficient generated by the self-learning module in the hot-rolled thickness control model, and can timely adjust the hot-rolled thickness control model, thereby improving the thickness control accuracy of the hot-rolled thickness control model in the hot-rolling process of the rolled piece.

The invention relates to the technical field of pile foundation construction quality monitoring, in particular to a real-time monitoring device and method for a pouring interface of an ultra-long cast-in-situ bored pile. Comprising a monitoring probe mechanism, a tension cable, a control holder, a display industrial personal computer and an alarm, the monitoring probe mechanism is placed in a bored hole of the cast-in-place pile, and the tension cable is electrically connected with the monitoring probe mechanism and the display industrial personal computer; the display industrial personal computer is located on the ground, and the monitoring probe mechanism uploads collected data to the display industrial personal computer on the ground in real time through a tension cable; high-precision real-time monitoring of the pouring interface of the super-long cast-in-situ bored pile is achieved, and it is guaranteed that the position of the pouring interface is accurate; meanwhile, the situation that the monitoring probe is blocked and is difficult to recycle is avoided, and the adaptive capacity of the whole device is greatly improved; continuous measurement of different pouring states and stages is achieved, the real-time monitoring device can feed back position information of a pouring interface in real time, constructors can adjust a construction scheme in time, and the construction quality is ensured.

The invention provides a method and device for assisting construction progress plan simulation calculation, and relates to the technical field of data processing, construction data of historical projects is collected, and the construction data comprises one or more of task information, process information, resource information and construction progress information; training a constructed prediction model by using the collected construction data to obtain a trained prediction model; and performing simulation calculation on the construction progress plan of the current project based on the trained prediction model, and judging whether delay occurs or not. Therefore, feature vectors of task information, process information, resource information, construction progress information and the like are extracted from the input actual construction progress data by using the deep learning model so as to predict a construction progress plan in real time and adjust the construction progress plan in time, and the effects of better allocating resources, reducing construction cost and improving benefits are achieved.

The invention relates to a diabetes exercise rehabilitation guidance system, relates to the technical field of diabetes rehabilitation, and aims to provide a personalized and scientific exercise management scheme for diabetics. The system comprises a user individual feature data acquisition module, an evaluation and classification module, an exercise scheme generation module, a real-time monitoring module and a feedback and adjustment module. By collecting the physiological parameters of the user in real time, the system automatically analyzes the health condition of the user and classifies the user into low-risk, medium-risk and high-risk exercise crowds. Based on a classification result, the system uses a machine learning algorithm to generate a personalized exercise scheme, and the safety of the exercise process is ensured through dynamic monitoring. The real-time monitoring module of the system can monitor the physiological data of the user in the exercise process and give an alarm in time when an abnormality is found. Through connection with an information system of a medical institution, the system also supports a remote medical function, so that doctors can check health data of patients in real time and provide professional guidance.

The invention discloses a financial data storage and query system, and relates to the technical field of electric digital data processing. The financial data storage and query system comprises a delay influence evaluation module, a data accessibility evaluation module and a data transmission security evaluation module. According to the invention, the delay influence value is obtained in the first preset time period and whether data accessibility evaluation is carried out is judged, and then the financial data accessibility evaluation value is obtained in the second preset time period and whether access adjustment is carried out is judged; and finally, obtaining the financial data secure transmission evaluation value in the third preset time period and judging whether secure transmission adjustment is performed or not, so that the effect of improving the timeliness in the financial data storage and query process is achieved, and the problem of low timeliness in the financial data storage and query process in the prior art is solved.

The invention provides a power transmission project GIM model analysis loading method and device, equipment and a medium, and belongs to the technical field of power transmission engineering. A power transmission project GIM model is obtained, and tower group graph data in the power transmission project GIM model is extracted and analyzed; performing simplification processing on the tower group graphic data according to different levels, and outputting a simplification result as a three-dimensional model; slicing the three-dimensional model to form a model file; extracting state information of tower equipment in the model file; analyzing the attribute information of the tower equipment in the power transmission project GIM model; carrying out three-dimensional space arrangement on tower equipment to form a form of a power transmission line; analyzing the positions of the grounding wire and the hanging grounding wire, and storing the key point data of the grounding wire; and based on the space range control model, setting a three-dimensional model, a loading range of a grounding wire and a hanging grounding wire and a visible height of the three-dimensional model. By setting the loading range and the visible height, the loading and rendering efficiency of the model can be optimized, and the interaction experience of the user is improved.

The invention discloses a recommendation optimization method based on multi-modal feature clustering, and aims to improve diversity and user experience of a recommendation system. According to the method, firstly, features are extracted from a commodity title and a cover picture through a text and visual feature extraction model, and text and visual feature representation is obtained; then, the features are subjected to feature alignment and standardization processing to ensure the consistency of different modal features, and then the features are fused into a multi-modal embedded vector to serve as unified feature representation of commodities. And then, clustering is carried out on the fused multi-modal embedded vectors by using a K-Means clustering algorithm, so that clustering identifiers of the commodities can be formed. In the recommendation process, the system monitors the clustering identifiers of the displayed commodities, adjusts the recommendation strategy according to the identifiers, and preferentially recommends the commodities of different clusters from the displayed commodities, so that excessive distribution of similar contents is avoided, and the diversity of the recommended contents is ensured. According to the method, by integrating multi-modal information, the performance of a recommendation system is optimized, and the problem of excessive distribution is effectively solved.

The invention relates to the field of intelligent management, data science and community services, in particular to a future community service intelligent management method and system based on big data, and the method comprises the core processes of data collection, feature extraction, intelligent agent task scheduling, resource optimization and the like. Feature data which can be used for analysis are generated through data fusion; analyzing a resource use state in real time by using a deep learning technology, and generating resource recommendation data; three-dimensional visual display is carried out through the holographic data management platform, and a decision maker can quickly identify problems and make adjustments; according to the invention, the efficiency and quality of community service are improved through intelligent regulation and control, and scientific basis and technical support are provided for community management.