An interface layout generation method based on an intelligent gateway and an intelligent gateway

Abstract

The application relates to an interface layout generation method and an intelligent gateway based on the intelligent gateway, relates to the technical field of Internet of Things, and the interface layout generation method comprises the following steps: acquiring and transmitting instrument monitoring video to the intelligent gateway, and identifying instrument readings; according to a device fault dynamic judgment mechanism combined with sensor data, instrument readings are fused and analyzed to obtain data analysis results, and corresponding operation trigger points are activated; according to the operation trigger points combined with a preset weight evaluation model, interface function weights are determined, a dynamic layout generation algorithm is combined to construct an interface layout template; user operation behaviors are captured and coded to generate interface operation events, the interface layout template is corrected in combination with event processing logic, and an optimal interface layout is obtained; the interface function weights are determined through the construction of the weight evaluation model, the dynamic layout generation algorithm and an interface layout template storage calling mechanism are combined, reasonable same-screen layouts are automatically generated according to different business requirements and use scenarios, and efficient and intuitive information display is obtained.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of Internet of Things, in particular to an interface layout generation method based on an intelligent gateway and the intelligent gateway. BACKGROUND

[0002] In the fields of electric power and industrial automation, a gateway device is a key node for data transmission and processing, and its function evolves from simple data forwarding to intelligent analysis and processing. Traditional gateways mainly focus on data transmission efficiency and stability, and although they have basic data transmission functions, they still have problems such as single function, low hardware integration, and non-uniform communication protocols, and they lack integrated intelligent analysis modules such as pointer instrument identification and digital instrument identification, and intuitive human-computer interaction interfaces for users.

[0003] Patent No. CN202410750395.0 discloses an interface layout automatic generation method and system, and a storage medium. In the existing vehicle-mounted intelligent system, the interface layout automatic generation method is loaded, the interface resolution parameters of the current interface are directly obtained, the style configuration parameters are calculated, and then the layout parameters of the basic grid in the current interface layout are calculated by combining the element parameters corresponding to the current interface style, so that the layout parameters of the current display interface are determined. At this time, the interface proportion of each application interface is allocated based on the current scene and / or user characteristics, and the interface partition is obtained by partitioning the basic grid. The application interface is laid out according to the user's demand. Then, the layout coordinates of each element in the application interface are further calculated, so that the layout and elements of the automatic generation UI picture are automatically generated according to the user's preference, the change of the use scene and the like, to provide the user with flexible and intelligent experience in line with the user's intention.

[0004] The prior art in the above has the following defects:

[0005] 1. The traditional gateway lacks intelligent analysis capability and cannot effectively identify pointer instrument and digital instrument information, thereby reducing the intelligent level of data processing.

[0006] 2. The existing technology has high failure rate and poor expansibility due to the dispersion of hardware interfaces, and lacks efficient and stable protocol support for communication with the superior master station. SUMMARY

[0007] In view of the deficiencies of the prior art, the application determines the interface function weight by constructing a weight evaluation model, and combines a dynamic layout generation algorithm and an interface layout template storage and calling mechanism to automatically generate a reasonable same-screen layout according to different business requirements and use scenes, and obtain efficient and intuitive information display.

[0008] The following technical solutions are adopted:

[0009] A smart gateway-based interface layout generation method, comprising:

[0010] Acquiring and transmitting instrument monitoring video to a smart gateway and identifying instrument readings;

[0011] According to a device fault dynamic judgment mechanism combined with sensor data, the instrument readings are fused and analyzed to obtain data analysis results, and corresponding operation trigger points are activated;

[0012] According to the operation trigger points combined with a preset weight evaluation model, interface function weights are determined, and a dynamic layout generation algorithm is used to construct an interface layout template;

[0013] User operation behavior is captured and coded to generate interface operation events, and the interface layout template is corrected combined with event processing logic to obtain an optimal interface layout.

[0014] By adopting the above technical solutions, the smart gateway and SIP-B protocol transmission technology are integrated to acquire instrument monitoring video, the pointer angle calculation and digital character recognition are realized by using filtering enhancement, edge detection and OCR algorithm, a multi-source feature fusion model is constructed combined with sensor data, machine learning is used to realize fault dynamic prediction and trigger operation instructions; based on the weight evaluation model, the interface business priority is quantified, the visual focus area is automatically distributed combined with screen parameters through the dynamic layout generation algorithm, the layout template library is established to realize scene adaptive switching; at the same time, the user behavior characteristics are captured by using the event listening mechanism, the operation response link is optimized through coding mapping and filtering rules, the "perception-analysis-decision- interaction" closed loop is formed combined with the animation feedback mechanism and the hierarchical data pushing strategy, the three goals of instrument reading accurate identification, fault real-time early warning and operation interface dynamic optimization are realized, and the method has the characteristics of high intelligent level, strong multi-protocol compatibility and efficient interface adaptive efficiency.

[0015] The application further provides that the specific steps of acquiring and transmitting instrument monitoring video to a smart gateway and identifying instrument readings comprise:

[0016] Monitoring all instruments to obtain initial instrument monitoring video;

[0017] Dividing the initial instrument monitoring video into blocks to obtain a plurality of instrument video blocks;

[0018] Performing permission authentication between the gateway host end and the main station receiving end to complete communication connection and build a transmission channel;

[0019] According to a preset redundant link protocol, all the instrument video blocks are encrypted and packaged to obtain a plurality of encrypted and packaged video blocks; the encrypted and packaged video blocks are transmitted to the main station receiving end through the transmission channel, and the data transmission process is monitored to obtain network status and transmission rate;

[0020] The transmission rate is corrected according to the network state, and the integrity of each encrypted and packaged video block is verified in combination with a cache retransmission mechanism;

[0021] If all the encrypted and packaged video blocks are complete and correct, the encrypted and packaged video blocks are decrypted, spliced according to timestamps, and an aggregated instrument video is obtained;

[0022] The aggregated instrument video is matched and frame-decomposed according to a preset instrument template, and an instrument time sequence frame is obtained;

[0023] The instrument time sequence frame is sampled and extracted according to a preset time interval, and an original instrument image is obtained;

[0024] The original instrument image is filtered and denoised, and a noise-free instrument image is obtained;

[0025] The noise-free instrument image is image-enhanced, and an enhanced instrument image is obtained;

[0026] The enhanced instrument image is classified according to an instrument type, and a pointer instrument image and a digital instrument image are obtained;

[0027] The pointer instrument image is edge-detected, and a dial outline and a pointer line are obtained;

[0028] The dial outline and the pointer line are fitted according to a preset angle calculation model in combination with an instrument reference scale, and an instrument reading is recognized;

[0029] The digital instrument image is subjected to affine transformation and edge detection, and a character display frame is obtained;

[0030] The character display frame is horizontally positioned and vertically divided according to pixel continuity, and a plurality of independent characters are obtained;

[0031] Each independent character is recognized and combined according to a preset character recognition model, and an instrument reading is obtained.

[0032] By adopting the above technical solution, the instrument video is collected by a monitoring camera in real time and is transmitted to a smart gateway in encrypted blocks, a transmission channel is constructed based on a redundant link protocol and is subjected to integrity verification and dynamic rate adjustment, the aggregated video is classified according to an instrument type by a YOLOv7 model, the pointer instrument reading is recognized by edge detection and angle fitting algorithms respectively, the digital instrument data is analyzed by an affine transformation and character segmentation model, the automatic and accurate collection and transmission of instrument data in an industrial scene are realized, and the solution has the advantages of efficient transmission, multi-protocol compatibility, dynamic anti-interference, and abnormal alarm, and can effectively reduce the cost of manual inspection and improve the efficiency of equipment state monitoring.

[0033] The application is further provided: the specific steps of activating the corresponding operation trigger point according to the device fault dynamic judgment mechanism combined with sensor data, fusion analysis of the instrument reading, obtaining data analysis results, include:

[0034] The running time of the equipment in different working condition modes is counted and arranged in descending order respectively to obtain a working condition working time sequence;

[0035] According to the time stamp, the working condition working time sequence is analyzed in association with the historical sensor data to obtain working condition association time sequence data; according to the equipment characteristics combined with the working condition association time sequence data, a static judgment threshold interval is determined;

[0036] The static judgment threshold intervals of different working condition modes are fitted and operated to obtain a full working condition dynamic function;

[0037] The instrument reading and sensor data are feature extracted to obtain instrument characteristic values and sensor characteristic values;

[0038] The instrument characteristic values and the sensor characteristic values are compared and judged with the static judgment threshold interval respectively; if any is not located in the static judgment threshold interval, it is determined that the equipment is suspected to be faulty, and the corresponding fault decision value is generated; if none is located in the static judgment threshold interval, it is determined that the equipment has failed;

[0039] According to the working condition mode combined with the full working condition dynamic function, a dynamic judgment threshold interval is generated, and is compared and judged with the instrument characteristic values and the sensor characteristic values respectively; if any is not located in the dynamic judgment threshold interval, the corresponding fault decision weight proportion coefficient is corrected; if none is located in the dynamic judgment threshold interval, the equipment failure is graded, and a fault level code is generated;

[0040] According to the fault decision weight proportion coefficient combined with the fault decision value, a fault trigger value is generated, and is matched with a preset fault level warning interval to obtain the fault level code;

[0041] According to the fault level code combined with the fault duration, the corresponding operation trigger point is activated.

[0042] By adopting the technical scheme, based on the fault diagnosis algorithm of time sequence aggregation analysis and dynamic threshold fitting, firstly, the working condition working time sequence is constructed by arranging the working condition mode running time in descending order, and the working condition associated time sequence data is generated by associating with the historical sensor data, the static threshold interval is extracted combined with the equipment characteristics, and the full working condition dynamic function is fitted; then the fault is determined by double comparison of instrument and sensor characteristic value and static / dynamic threshold (static threshold triggers preliminary fault decision value, dynamic threshold corrects decision weight), finally, the fault level code is generated by fusing the weight coefficient and the fault level warning interval, and the corresponding operation trigger point is activated, the advantages are that the time sequence clustering, multi-stage statistical modeling, static time sequence anomaly detection and dynamic function fitting technology are fused, through the double determination mechanism of static threshold preliminary screening and dynamic threshold adaptive calibration, the fault recognition accuracy is significantly improved and the false alarm rate in the multi-working condition scene is reduced, at the same time, the adaptability of the system to nonlinear working condition change is strengthened relying on the time sequence data association and weight correction strategy, the intelligent closed-loop control of fault grading response and resource scheduling is realized.

[0043] The application further provides that: the specific steps of constructing the interface layout template according to the operation trigger point combined with the preset weight evaluation model, including:

[0044] According to the user operation frequency and data update frequency, the operation interface is divided, and the high-frequency touch area is marked; according to the business process, the high-frequency touch area is disassembled, the key trigger action is marked, and the function interface area is divided;

[0045] According to the business value and information priority, the function interface area is constructed to obtain a function trigger matrix;

[0046] According to the interface size, a responsive rule is set, and the function trigger matrix is iteratively trained combined with the operation trigger point to obtain a weight evaluation model, and an interface function weight is output;

[0047] According to the interface function weight combined with the dynamic layout generation algorithm, all function areas are allocated to obtain a corresponding interface space proportion;

[0048] According to the interface space proportion combined with the Gestalt principle, all the function areas are clustered and deployed to obtain an interface layout template.

[0049] By adopting the above technical solutions, high-frequency touch areas are divided using the K-means clustering algorithm of heatmaps, a function trigger matrix is ​​established by decomposing business processes, business value indicators are constructed using the analytic hierarchy process, a weight evaluation model is generated by iterative training of a BP neural network under responsive rules, the interface space ratio is allocated based on a dynamic layout genetic algorithm, and finally, the interface layout template is realized through Gestalt-based perceptual clustering, forming a data-driven adaptive interface optimization system. Its advantage lies in dynamically adjusting the layout weights through machine learning, combining user behavior analysis and cognitive psychology principles to achieve spatial matching between interface functional areas and operating habits. While ensuring business process efficiency, it significantly improves visual hierarchy perception, enhances the efficiency of high-frequency function triggering, and has cross-terminal response capabilities.

[0050] The present invention is further configured to: set responsive rules according to the interface size, and iteratively train the function trigger matrix in combination with operation trigger points to obtain a weight evaluation model, and output the interface function weights. The specific steps include:

[0051] The user interface is divided into preset size groups according to the interface size, and the number of grid columns and component priorities are defined.

[0052] Based on the number of grid columns and the component priority, combined with layout adaptation logic, responsive rules are generated;

[0053] Based on the aforementioned responsive rules and the operation trigger points, weights are assigned to the function trigger matrix to form a weight vector matrix;

[0054] Based on the preset constraint parameters and the weight vector matrix, the function trigger matrix is ​​trained iteratively several times to obtain the weight evaluation model.

[0055] The initial weight matrix W of the weight evaluation model is optimized based on the target loss function L to obtain the optimal weight matrix;

[0056]

[0057] Where j is the interface size, n is the maximum interface size, i is the functional module number, k is the total number of functional modules, and w i For the global weight of the functional module, s j λ is the device size weighting coefficient, M[i,j] is the element of the function trigger matrix, λ is the regularization coefficient, and Reg is the regularization term;

[0058] The user interface is weighted according to the optimal weight matrix and user behavior data, and the interface function weights are output.

[0059] By adopting the technical scheme, the number of columns and the component priority are defined based on the preset size group division interface and the grid dynamic planning algorithm, the function trigger matrix is constructed in combination with the responsive breakpoint rule and the operation heat map data, the L2 regularization optimization objective function is constrained by introducing the device size weight coefficient and the regularization term, the weight evaluation model is iteratively trained by using the gradient descent algorithm, and finally the optimal weight matrix is generated by dynamically adjusting the layout by using the genetic algorithm, so that the multi-terminal adaptive interface function weight distribution is realized; The advantage is that the model complexity is controlled by regularization to avoid overfitting, the space proportion is dynamically optimized in combination with the device characteristics and user behavior, the high-frequency function layout and visual motion line are matched in the cross-resolution scene, the interface operation efficiency is improved, and the multi-terminal adaptation cost is reduced based on the grid elasticity.

[0060] The application further provides that: the specific steps of capturing and encoding the user operation behavior, generating the interface operation event, and correcting the interface layout template in combination with the event processing logic to obtain the optimal interface layout include:

[0061] The user operation behavior on the operation interface is captured and counted to obtain the interface operation event;

[0062] All the interface operation events are filtered for invalid operations according to the event listening mechanism to obtain valid operation events;

[0063] The valid operation events are classified and encoded to obtain unique event codes;

[0064] The function areas are sorted according to the unique event codes in combination with the event processing logic to generate a function trigger process;

[0065] The function trigger process is disassembled and associated with the function modules of the device to obtain a process mapping relationship table; and the interface layout template is corrected according to the process mapping relationship table in combination with the user operation habits to obtain the optimal interface layout.

[0066] By adopting the technical scheme, the user operation behavior is captured by the event listening mechanism and filtered for invalid operations by using the dynamic threshold algorithm, the valid operation events are classified and encoded to generate unique identifiers by using the clustering algorithm, the function trigger process is constructed by using the hidden Markov model (HMM), and the process mapping relationship table is constructed based on the association rule mining, and finally the interface layout is dynamically optimized by using the genetic algorithm or the reinforcement learning model, so that the interface is sorted and adjusted according to the user operation habits, and the advantages are that the data-driven closed-loop optimization mechanism reduces the user cognitive load while improving the operation efficiency, and has the real-time response and personalized adaptation capability.

[0067] In the second aspect, the application further provides an interface layout generation system based on an intelligent gateway, which adopts the following technical scheme:

[0068] An interface layout generation system based on a smart gateway, comprising:

[0069] A sensor monitoring module for real-time display of device deployment environment parameter data collected by sensors;

[0070] An AI recognition module for analyzing and recognizing pointer instruments and digital instruments through image recognition technology to obtain instrument readings;

[0071] A plane layout module for visualizing the physical location and connection topology relationship of devices;

[0072] A state monitoring module for monitoring the running state and abnormal alarm information of the smart gateway;

[0073] A video monitoring module for real-time display of station site video stream and remote instrument monitoring pictures;

[0074] An intelligent inspection module for automatically executing inspection tasks and feeding back device running state results;

[0075] An AI analysis module for fusion of multi-source data to realize fault prediction and intelligent decision analysis;

[0076] A linkage control module for triggering device control and emergency response operation according to preset logic;

[0077] A historical query module for storing and tracing back historical data and statistical analysis results.

[0078] By adopting the above technical solution, the sensor monitoring module collects environmental parameters (such as temperature and humidity, illumination, etc.) in real time, the AI recognition module analyzes instrument images based on a convolutional neural network (CNN) algorithm to obtain accurate readings, the plane layout module visualizes device location and connection relationship by using a topology optimization algorithm, the state monitoring module diagnoses gateway running state in real time by using a time series anomaly detection algorithm and triggers alarms, the video monitoring module synchronously presents station live pictures by using a multi-target tracking algorithm, the intelligent inspection module autonomously executes inspection tasks by using a path planning algorithm, the AI analysis module realizes multi-source data driven fault prediction and decision optimization by using a machine learning model (such as LSTM, random forest), the linkage control module dynamically adjusts device running strategy by using a rule engine, and the historical query module traces back data trends by using a time series database and a visualization algorithm. Ultimately, an intelligent management system integrating monitoring, recognition, analysis, and control is constructed, management efficiency and decision accuracy are greatly improved, the need for manual intervention is reduced, system real-time performance and reliability are enhanced, full life cycle support is provided for device operation and maintenance, multi-modal data fusion and adaptive logic optimization resource allocation are realized, fault early warning accuracy is improved, and inspection efficiency is improved.

[0079] Thirdly, the present invention also provides an electronic device, which adopts the following technical solution:

[0080] An electronic device, comprising:

[0081] One or more processors;

[0082] Memory, used to store one or more programs;

[0083] When the one or more programs are executed by the one or more processors, the one or more processors implement the method described above.

[0084] By adopting the above technical solution, the interface layout generation method based on the smart gateway is presented in the form of computer-readable code and stored in the memory. When the processor runs the computer-readable code in the memory, the steps of the interface layout generation method based on the smart gateway are executed, thereby reducing the intensity of manual labor and improving the degree of automation.

[0085] Fourthly, the present invention also provides a computer storage medium, which adopts the following technical solution:

[0086] A computer storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0087] In summary, the beneficial technical effects of the present invention are as follows:

[0088] 1. By constructing a weight evaluation model to determine the weight of interface functions, and combining a dynamic layout generation algorithm and a layout template storage and retrieval mechanism, it can automatically generate reasonable on-screen layouts according to different business needs and usage scenarios, resulting in efficient and intuitive information display.

[0089] 2. Classify and encode operation events, establish mapping relationships with functional modules, and combine with real-time capture and filtering mechanisms to ensure that the system can accurately identify user operation intentions and quickly call the corresponding functional modules for processing.

[0090] 3. Based on data update-triggered monitoring, prioritize updated information and adopt a tiered push and display strategy to enable devices to intelligently process information of different importance. Attached Figure Description

[0091] Figure 1 This is a flowchart illustrating an interface layout generation method according to one embodiment of the present invention.

[0092] Figure 2 This is a flowchart illustrating an interface layout generation method according to one embodiment of the present invention.

[0093] Figure 3 is a flowchart of an interface layout generation method according to an embodiment of the present application.

[0094] Figure 4 is a schematic diagram of an interface layout generation system according to an embodiment of the present application. DETAILED DESCRIPTION

[0095] The present application will be further described in detail below with reference to the accompanying drawings.

[0096] Reference Figure 1 A method for generating an interface layout based on an intelligent gateway is disclosed, comprising:

[0097] S1: acquiring and transmitting instrument monitoring video to an intelligent gateway, and identifying instrument readings;

[0098] S2: combining sensor data according to a device fault dynamic judgment mechanism, fusing and analyzing the instrument readings to obtain a data analysis result, and activating a corresponding operation trigger point;

[0099] S3: determining an interface function weight according to the operation trigger point combined with a preset weight evaluation model, and constructing an interface layout template combined with a dynamic layout generation algorithm;

[0100] S4: capturing and encoding user operation behavior, generating an interface operation event, and correcting the interface layout template combined with event processing logic to obtain an optimal interface layout.

[0101] The implementation principle of the embodiment is: receiving multiple instrument monitoring video streams through an intelligent gateway, using a dynamic code rate adjustment algorithm to ensure transmission stability, and identifying readings based on an improved YOLOv7 visual model; synchronously accessing multi-frequency sensor data, implementing multi-source information fusion through timestamp alignment and Kalman filtering, and generating dynamic operation trigger points combined with a fault knowledge graph and a Bayesian network; a weight evaluation model analyzes user historical behavior data through a reinforcement learning module, dynamically optimizes function priority, and drives a genetic algorithm to generate an initial interface layout; after user operation behavior is encoded in real time through an edge node, an event-driven engine is triggered, layout parameters are iteratively corrected combined with operation success rate, delay, and preference tags, and finally an optimal interface adapted to a terminal is output through distributed rendering services.

[0102] Embodiment Two

[0103] The specific steps of step S1 include:

[0104] Monitoring all instruments to obtain initial instrument monitoring video;

[0105] Segmenting the initial instrument monitoring video to obtain a plurality of instrument video blocks;

[0106] The gateway host end and the main station receiving end are authenticated for authority, a communication connection is completed, and a transmission channel is constructed.

[0107] According to a preset redundancy link protocol, all the instrument video blocks are encrypted and packaged to obtain a plurality of encrypted and packaged video blocks; the encrypted and packaged video blocks are transmitted to the main station receiving end through the transmission channel, and a data transmission process is monitored to obtain a network state and a transmission rate;

[0108] The transmission rate is corrected according to the network state, and each encrypted and packaged video block is integrity-verified in combination with a cache retransmission mechanism.

[0109] In this embodiment, the video information is packaged by the gateway host end according to the SIP-B protocol, and the main station receiving end performs analysis and processing; a handshake protocol is used to establish a connection and authenticate authority, a connection state is monitored in real time, and automatic reconnection is used to ensure transmission continuity; the transmission rate and the encoding format are dynamically adjusted according to the network condition, and the cache and the retransmission mechanism are used to ensure data integrity and accuracy.

[0110] In order to solve the problem of insufficient support of the gateway in compatibility with multiple communication protocols, especially the SIP-B protocol, a highly modular design idea is adopted, and multiple interface resources are integrated, including POE network, storage, serial port, small current sampling, input and output signal control, 4G / 5G communication, LORA, WIFI, Bluetooth, etc.

[0111] If all the encrypted and packaged video blocks are complete and correct, the encrypted and packaged video blocks are decrypted, and spliced according to the time stamp to obtain an aggregated instrument video.

[0112] According to a preset instrument template, the aggregated instrument video is matched and frame-decomposed to obtain an instrument time sequence frame.

[0113] According to a preset time interval, the instrument time sequence frame is sampled and extracted to obtain an original instrument image.

[0114] The original instrument image is filtered and denoised to obtain a noise-free instrument image.

[0115] The noise-free instrument image is image-enhanced to obtain an enhanced instrument image.

[0116] According to the instrument type, the enhanced instrument image is classified to obtain a pointer instrument image and a digital instrument image.

[0117] The pointer instrument image is edge-detected to obtain a dial outline and a pointer line.

[0118] According to a preset angle calculation model in combination with an instrument reference scale, the dial outline and the pointer line are fitted to identify an instrument reading.

[0119] affine transform and edge detection are performed on the digital instrument image to obtain a character display frame;

[0120] According to the pixel continuity, the character display frame is horizontally positioned and vertically divided to obtain a plurality of independent characters;

[0121] According to a preset character recognition model, each independent character is recognized and combined to obtain an instrument reading.

[0122] In this embodiment, the instrument image is collected by the built-in image acquisition module, and the image quality is improved through filtering and enhancement algorithm processing. The dial contour and pointer line are extracted through edge detection, the pointer value is determined by combining the angle calculation model, and the scale calibration is performed. The OCR technology is used for character segmentation, feature extraction and pattern matching, and the character template library is established and optimized.

[0123] The implementation principle of this embodiment is: align the timestamps of multi-channel instrument video and sensor data through SIP-B protocol, use SRT protocol to superimpose quantum key for double-link encryption transmission of block video stream, combine MEMS inertial sensor to dynamically compensate image distortion caused by mechanical vibration; verify data integrity based on blockchain hash chain during transmission, trigger local cache and TEE encryption retransmission of edge node in abnormal condition; the receiving end drives the improved CRNN model to recognize the pointer / digital instrument reading by using the anti-interference character library generated by multi-spectral fusion enhancement and GAN; at the same time, dynamically adjust the video code rate and encryption strength according to the network state, and finally synchronize the aggregated video and high-confidence reading data to the knowledge graph through space-time consistency verification.

[0124] Embodiment three:

[0125] The specific steps of step S2 include:

[0126] The running time of the equipment in different working condition modes is counted and arranged in descending order to obtain a working condition working time sequence.

[0127] According to the time stamp, the working condition working time sequence and historical sensor data are associated and analyzed to obtain working condition associated time sequence data; in this embodiment, the priority of the main sensor and the standby sensor can also be set, and the initial working mode and data acquisition frequency are defined. In the initialization stage, the normal working range, error allowable value and fault determination standard (such as threshold value of data fluctuation amplitude, abnormal duration, etc.) of each sensor need to be determined.

[0128] According to the characteristics of the equipment and the working condition associated time sequence data, a static judgment threshold interval is determined;

[0129] The static judgment threshold intervals of different working condition modes are fitted and operated to obtain a full-working-condition dynamic function;

[0130] In this embodiment, according to historical data and equipment characteristics, a preset static threshold (such as a temperature sensor ± 5 ℃ deviation) or a dynamic threshold (such as a sliding window statistical value based on a time series) is calibrated to adapt to different working conditions.

[0131] Feature extraction is performed on the instrument readings and sensor data to obtain instrument feature values and sensor feature values;

[0132] The instrument feature values and the sensor feature values are compared with the static judgment threshold interval respectively; if any is not located in the static judgment threshold interval, it is determined that the equipment is suspected to be faulty, and a corresponding fault decision value is generated; if none is located in the static judgment threshold interval, it is determined that the equipment has failed;

[0133] In this embodiment, the output data of each sensor is monitored in real time, and compared with a preset threshold range. If the data continuously exceeds the threshold (such as exceeding the limit for 3 sampling periods), a primary fault alarm is triggered, and the sensor is marked as “suspicious state”.

[0134] According to the working condition mode and the full-condition dynamic function, a dynamic judgment threshold interval is generated, and is compared with the instrument feature values and the sensor feature values respectively; if any is not located in the dynamic judgment threshold interval, the corresponding fault decision weight proportion coefficient is corrected; if none is located in the dynamic judgment threshold interval, the equipment failure is graded, and a fault level code is generated;

[0135] In this embodiment, the working condition mode includes normal operation (parameters are stable within the design range, and the equipment works efficiently), attention state (parameters approach the limit but do not exceed the limit, and need to be monitored), abnormal state (important parameters approach or slightly exceed the limit, and need to be arranged for maintenance), and serious state (key parameters seriously exceed the limit, and must be immediately shut down for maintenance).

[0136] In this embodiment, the spatial consistency of abnormal data can also be identified through cross verification of redundant sensor data. For example, if the main sensor temperature value is abnormal but the backup sensor data is normal, it is determined that the main sensor is faulty; if both are abnormal, it may be a real environmental change or a system-level failure.

[0137] According to the fault decision weight proportion coefficient and the fault decision value, a fault trigger value is generated, and is matched with a preset fault level warning interval to obtain the fault level code;

[0138] In this embodiment, the fault level is also divided according to the deviation threshold degree and the duration (such as slight drift and serious failure). For example:

[0139] First-level fault: the data fluctuation is within 10% of the threshold, triggering a warning but not switching the equipment;

[0140] Secondary fault: 20% above threshold and lasts for 5 seconds, triggers primary backup switching;

[0141] Tertiary fault: multiple sensors abnormal at the same time, triggers system emergency shutdown.

[0142] According to the preset priority, automatically switch to the backup sensor, update the control right and feedback the switching state to the upper computer. The switching process needs to ensure the continuity of the control signal (such as using a smooth transition algorithm) to avoid system shock.

[0143] According to the fault level code combined with the fault duration, activate the corresponding operation trigger point.

[0144] The implementation principle of the embodiment is: through the controller to capture the equipment electrical signal jump point in real time, combined with the edge computing node to clean up the timing data and record the working condition start and stop time; based on the LSTM model to analyze the historical sensor data to generate a dynamic threshold function, and simultaneously use Kalman filter to fuse the main and backup sensor data to eliminate local anomalies; when the instrument characteristic value deviates from the threshold, through multi-spectrum verification and spatial consistency analysis to distinguish sensor failure or equipment abnormality, and according to the fault level code to trigger MES work order distribution and visual alarm screen alarm; the main and backup switching process is embedded with PID smoothing algorithm to ensure control continuity, at the same time based on blockchain to store fault decision records to realize maintenance traceability, finally through the adaptive learning engine to iteratively optimize the threshold interval, forming an intelligent monitoring system of equipment state perception-dynamic threshold judgment-multi-level fault response-closed loop knowledge update.

[0145] Embodiment four:

[0146] Referring to Figure 2 , the specific steps of step S3 include:

[0147] S31: According to the user operation frequency and data update frequency combined with the heat map, the operation interface is divided, and the high-frequency touch area is marked;

[0148] S32: According to the business process, the high-frequency touch area is disassembled, the key trigger action is marked, and the function interface area is divided;

[0149] In this embodiment, the key trigger action includes button click, form submission; the function interface area includes main function area (such as payment), secondary function area (such as help entry) and extension reserved area (such as sharing);

[0150] S33: According to the business value and information priority, the function interface area is constructed to obtain a function trigger matrix; S34: According to the interface size, set the responsive rule, and combine the operation trigger point to iteratively train the function trigger matrix to obtain a weight evaluation model, and output the interface function weight;

[0151] S35: According to the interface function weight and the dynamic layout generation algorithm, all function areas are allocated to obtain the corresponding interface space proportion;

[0152] In this embodiment, the interface space proportions are as follows: the main function area occupies 40%-60% of the screen, a fixed grid layout (such as 12 columns of Bootstrap system) is used, the secondary function area occupies 20%-30%, a flexible layout (Flexbox) is used to adapt to dynamic content, the extension reserved area reserves 10%-15% space margin, and subsequent function iteration is supported.

[0153] S36: According to the interface space proportion and the Gestalt principle, all the function areas are clustered and deployed to obtain an interface layout template.

[0154] The implementation principle of this embodiment is: by fusing multi-dimensional user behavior data (including space-time distribution heat map and scenario operation flow) and interactive physical rules, a dynamic function trigger matrix is constructed, combined with responsive device adaptation and personalized portrait, an initial layout is generated by using a Gestalt visual clustering algorithm; then based on real-time operation data backflow, the function weight is iteratively optimized by an online learning model, the flexible layout engine allocates space according to the main and secondary function area proportion, the A / B test module is embedded to verify the visual dynamic line and operation efficiency, and finally an adaptive interface system is formed, which takes into account business objectives, user experience and iterative expandability. In combination of statistical laws, cognitive psychology principles and machine learning dynamic parameter adjustment, a closed-loop optimization of “data perception-rule reasoning-layout generation-effect verification” is realized, and it is ensured that the interface space allocation always matches the user's high-frequency demand and the business evolution direction.

[0155] Embodiment five

[0156] Referring to Figure 3 , the specific steps of step S34 include:

[0157] S341: The operation interface is divided into a preset size group according to the interface size, and the grid column number and component priority are defined;

[0158] In this embodiment, the preset size group includes mobile terminal <768px, tablet 768-1024px, and desktop terminal >1024px, and layout rules (such as grid column number and component priority) are defined for each group;

[0159] S342: According to the grid column number and the component priority, a responsive rule is generated by combining layout adaptation logic;

[0160] S343: According to the response rule, the weight is assigned to the function trigger matrix combined with the operation trigger point to form a weight vector matrix; S344: According to the preset constraint parameter combined with the weight vector matrix, the function trigger matrix is iteratively trained several times to obtain a weight evaluation model;

[0161] S345: According to the target loss function L, the initial weight matrix W of the weight evaluation model is optimized to obtain an optimal weight matrix;

[0162]

[0163] Wherein, j is the interface size, n is the maximum interface size, i is the function module number, k is the total number of function modules, w i is the global weight of the function module, s j is the device size weight coefficient, M[i,j] is the function trigger matrix element, λ is the regularization coefficient, Reg is the regularization term;

[0164] S346: According to the optimal weight matrix combined with user behavior data, the operation interface is weighted and distributed, and the interface function weight is output.

[0165] The implementation principle of the embodiment is: a dynamic function trigger matrix is constructed by fusing multi-dimensional user behavior data (including space-time distribution heat map, device input modal feature) and interaction physical law, a differentiated response rule is generated based on device identification module and scene state machine, and personalized weight distribution is realized combined with decay factor; The target loss function embedded with visual inertia term is used to drive the weight evaluation model to iterate, and the matrix parameters are optimized by simultaneously absorbing real-time operation data through online learning mechanism, and finally relying on the elastic layout engine of Gestalt continuity constraint, the self-adaptive interface meeting the business target, interaction efficiency and cognitive consistency is output.

[0166] Embodiment six

[0167] The specific steps of step S4 include:

[0168] The user operation behavior on the operation interface is captured and counted to obtain interface operation events;

[0169] According to the event listening mechanism, all the interface operation events are filtered for invalid operations to obtain valid operation events;

[0170] The valid operation events are classified and coded to obtain unique event codes;

[0171] In this embodiment, operation events are classified and coded, invalid events are captured and filtered in real time by using a listening mechanism, function modules are called for processing through a mapping relationship, operation feedback is provided through animation and sound, trigger points are set, update information priority is divided, hierarchical pushing and display strategies are adopted, and user self-defined settings are supported.

[0172] According to the unique event code combined with event processing logic, the function area is sorted to generate a function trigger flow;

[0173] The function trigger flow is disassembled and associated with the function modules of the device to obtain a flow mapping relationship table; according to the flow mapping relationship table combined with user operation habits, the interface layout template is corrected to obtain the optimal interface layout.

[0174] The implementation principle of the embodiment is: based on a dynamic threshold algorithm of a time window and trajectory curvature, and integrating system-level permissions to realize global touch event listening; through global touch event listening and dynamic threshold filtering to capture effective operations, based on standardized coding rules and asynchronous event queues to realize high-concurrency processing; combined with device input modality recognition layer to adapt to multi-end interaction differences, using sliding time window and LSTM model to dynamically update user habit weights to drive function flow mapping optimization; through a composite feedback channel and transactional log to enhance interaction fault tolerance, and finally relying on thread priority scheduling and breakpoint resume strategy to output an adaptive interface that takes into account real-time response, cross-end compatibility and behavior prediction accuracy, to realize dynamic optimal matching of user intent and interface response.

[0175] Embodiment seven:

[0176] Referring to Figure 4 , an interface layout generation system based on an intelligent gateway, comprising:

[0177] A sensor monitoring module for real-time display of device deployment environment parameter data collected by sensors;

[0178] An AI recognition module for analyzing and identifying pointer instruments and digital instruments through image recognition technology to obtain instrument readings;

[0179] A planar arrangement module for visualizing the physical location and connection topology relationship of the device;

[0180] A state monitoring module for monitoring the running state and abnormal alarm information of the intelligent gateway;

[0181] A video monitoring module for real-time display of station site video stream and remote instrument monitoring pictures;

[0182] An intelligent inspection module for automatically executing inspection tasks and feeding back device running state results;

[0183] An AI analysis module is configured to fuse multi-source data to realize fault prediction and intelligent decision analysis.

[0184] A linkage control module is configured to trigger device control and emergency response operation according to preset logic.

[0185] A historical query module is configured to store and trace historical data and statistical analysis results.

[0186] The implementation principle of the embodiment is as follows: first, a weight evaluation model is constructed based on the business importance, data update frequency and operation frequency of nine modules such as sensor monitoring and AI recognition, and the interface function weight is determined. According to the interface function weight, a dynamic layout generation algorithm is designed. High-weight interfaces are allocated in the central part of the screen or in the visual key area, and low-weight interfaces are placed at the edge or processed by folding, hiding and the like. A layout template library is established, and the layout schemes generated under different weight combinations and screen parameters are stored. When the system starts or the user switches the use scene, the corresponding layout template is quickly called according to the current interface weight and device parameters, so as to realize efficient initialization and switching of the operation interface.

[0187] The user's operations on the operation interface, such as clicking, sliding, long pressing, double-finger zooming and the like, are classified and assigned with unique event codes. A mapping relationship table of operation events and function modules is established, and the system functions and processing logic corresponding to each operation event are determined. The event listening mechanism of the operating system is used to capture the user's operation behavior in real time. Event filtering rules are set to filter repeated and invalid operation events, so as to avoid waste of system resources. For example, operations of clicking the same area multiple times in a short time only keep the first valid operation. When the system receives a valid operation event, the corresponding function module is called for processing according to the event code and the mapping relationship table.

[0188] In the data acquisition module and the system state monitoring module of the gateway device, data update trigger points are set. When new data is collected by the sensor, the instrument recognition result changes or the system state is abnormal, a data update event is triggered, and the type, source and other information of the updated data are encapsulated. According to the importance and urgency of the data, the update information is prioritized. For high-priority information, a strong reminder method such as full-screen pop-up window and high-frequency flashing is used for pushing, and the information is displayed in a prominent position of the operation interface; for low-priority information, a light-weight method such as notification bar prompt and icon badge is used for pushing, and the data is updated in the corresponding interface.

[0189] An electronic device comprises:

[0190] one or more processors;

[0191] a memory for storing one or more programs;

[0192] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as above.

[0193] A computer storage medium having stored thereon a computer program, which program, when executed by a processor, implements the method as above.

[0194] The embodiments of the present disclosure are all the preferred embodiments of the present disclosure, and are not intended to limit the protection scope of the present disclosure, so that: any equivalent changes made according to the structure, shape, principle of the present disclosure should be covered within the protection scope of the present disclosure.

Claims

1. A method for generating interface layout based on a smart gateway, characterized in that, include: Acquire and transmit instrument monitoring video to the smart gateway, and identify instrument readings; Based on the equipment fault dynamic judgment mechanism and sensor data, the instrument readings are fused and analyzed to obtain data analysis results, and corresponding operation trigger points are activated, including: The operating time of the equipment under different operating conditions is statistically analyzed and sorted in descending order to obtain the operating time sequence. Based on the timestamp, the working time sequence of the described working condition is correlated with historical sensor data to obtain working condition correlated time sequence data; Based on the equipment characteristics and the associated time-series data of the operating conditions, determine the static judgment threshold range; The static judgment threshold ranges of different working conditions are fitted to obtain the dynamic function for all working conditions. Feature extraction is performed on instrument readings and sensor data to obtain instrument feature values ​​and sensor feature values; The instrument feature value and the sensor feature value are compared with the static judgment threshold interval respectively; if either is not within the static judgment threshold interval, the equipment is suspected to be faulty and a corresponding fault decision value is generated; if neither is within the static judgment threshold interval, the equipment is determined to be faulty. Based on the operating mode and the full-condition dynamic function, a dynamic judgment threshold range is generated and compared with the instrument feature value and the sensor feature value respectively. If any of them is not within the dynamic judgment threshold range, the corresponding fault decision weight ratio coefficient is adjusted. If none of them are within the dynamic judgment threshold range, the equipment fault is judged and a fault level code is generated. Based on the fault decision weight ratio coefficient and the fault decision value, a fault trigger value is generated and matched with a preset fault level warning range to obtain the fault level code. Based on the fault level code and the fault duration, activate the corresponding operation trigger point; Based on the operation trigger point and a preset weight evaluation model, the interface function weights are determined, and an interface layout template is constructed using a dynamic layout generation algorithm, including: Based on the user operation frequency and data update frequency combined with the heat map, the user interface is divided and high-frequency touch areas are marked. Capture and encode user actions, generate interface operation events, and combine event handling logic to modify the interface layout template to obtain the optimal interface layout.

2. The interface layout generation method based on a smart gateway according to claim 1, characterized in that, The specific steps for acquiring and transmitting instrument monitoring video to the smart gateway and identifying instrument readings include: All instruments are monitored to obtain initial instrument monitoring videos; The initial instrument monitoring video is divided into blocks to obtain several instrument video blocks; Perform authentication between the gateway host and the main station receiver to complete the communication connection and build a transmission channel; All the instrument video blocks are encrypted and encapsulated according to a preset redundant link protocol to obtain several encrypted and encapsulated video blocks. The encrypted and encapsulated video block is transmitted to the main station receiving end through the transmission channel, and the data transmission process is monitored to obtain the network status and transmission rate. The transmission rate is adjusted according to the network status, and the integrity of each encrypted video block is verified by combining the buffer retransmission mechanism. If all the encrypted encapsulated video blocks are complete and error-free, the encrypted encapsulated video blocks are decrypted and spliced ​​together according to the timestamps to obtain the aggregated instrument video.

3. The interface layout generation method based on a smart gateway according to claim 2, characterized in that, The specific steps for acquiring and transmitting instrument monitoring video to the smart gateway and identifying instrument readings also include: The aggregated instrument video is matched and decomposed into frames according to the preset instrument template to obtain the instrument time sequence frame; The instrument timing frames are sampled and extracted according to a preset time interval to obtain the original instrument image; The original instrument image is filtered and denoised to obtain a noise-free instrument image; The noise-free instrument image is enhanced to obtain an enhanced instrument image; The enhanced instrument images are classified according to the instrument type to obtain pointer instrument images and digital instrument images; Edge detection is performed on the pointer instrument image to obtain the dial outline and pointer lines; Based on a preset angle calculation model and the instrument's reference scale, the dial outline and the pointer lines are fitted to identify the instrument reading; Perform affine transformation and edge detection on the digital instrument image to obtain the character display border; Based on pixel continuity, the character display border is horizontally positioned and vertically divided to obtain several independent characters; Each individual character is identified and combined according to a preset character recognition model to obtain the instrument reading.

4. The interface layout generation method based on a smart gateway according to claim 1, characterized in that, The specific steps of determining the interface function weights based on the operation trigger points and a preset weight evaluation model, and constructing the interface layout template using a dynamic layout generation algorithm, further include: Based on the business process, high-frequency touch areas are broken down, key trigger actions are marked, and functional interface areas are divided. Based on business value and information priority, indicators are constructed for the functional interface areas to obtain a functional trigger matrix; The responsive rules are set according to the interface size, and the function trigger matrix is ​​iteratively trained in combination with the operation trigger points to obtain the weight evaluation model and output the interface function weights. Based on the interface function weights and the dynamic layout generation algorithm, all functional areas are allocated to obtain the corresponding interface space proportions. Based on the interface space ratio and Gestalt principles, all functional areas are clustered and deployed to obtain an interface layout template.

5. The interface layout generation method based on a smart gateway according to claim 4, characterized in that, The specific steps for setting responsive rules based on interface size, iteratively training the function trigger matrix in conjunction with operation trigger points to obtain a weight evaluation model, and outputting interface function weights include: The user interface is divided into preset size groups according to the interface size, and the number of grid columns and component priorities are defined. Based on the number of grid columns and the component priority, combined with layout adaptation logic, responsive rules are generated; Based on the aforementioned responsive rules and the operation trigger points, weights are assigned to the function trigger matrix to form a weight vector matrix; Based on the preset constraint parameters and the weight vector matrix, the function trigger matrix is ​​trained iteratively several times to obtain the weight evaluation model. The initial weight matrix W of the weight evaluation model is optimized based on the target loss function L to obtain the optimal weight matrix; ； Where j is the interface size, n is the maximum interface size, i is the functional module number, k is the total number of functional modules, and w i For the global weight of the functional module, s j λ is the device size weighting coefficient, m[i,j] is the element of the function trigger matrix, λ is the regularization coefficient, and Reg is the regularization term; The user interface is weighted according to the optimal weight matrix and user behavior data, and the interface function weights are output.

6. The interface layout generation method based on a smart gateway according to claim 1, characterized in that, The specific steps for capturing and encoding user actions, generating interface operation events, and modifying the interface layout template based on event handling logic to obtain the optimal interface layout include: Capture and analyze user actions on the user interface to obtain interface operation events; The event listening mechanism is used to filter out invalid operations from all the interface operation events to obtain valid operation events. The valid operation events are classified and coded to obtain unique event codes; Based on the unique event code and the event processing logic, the functional areas are sorted to generate a function triggering process; The function triggering process is broken down and associated with the functional modules of the device to obtain a process mapping relationship table; Based on the process mapping table and user operating habits, the interface layout template is modified to obtain the optimal interface layout.

7. A smart gateway, characterized in that, include: The sensor monitoring module is used to display the device deployment environment parameter data collected by the sensors in real time; The AI ​​recognition module is used to analyze and recognize pointer and digital instruments through image recognition technology to obtain instrument readings; The floor plan module is used to visualize the physical location and connection topology of the devices; The status monitoring module is used to monitor the operating status and abnormal alarm information of the smart gateway; The video monitoring module is used to display real-time video streams from the station site and remote instrument monitoring screens; The intelligent inspection module is used to automatically perform inspection tasks and provide feedback on the equipment operating status. The AI ​​analysis module is used to integrate multi-source data to achieve fault prediction and intelligent decision analysis. The linkage control module is used to trigger equipment control and emergency response operations according to preset logic; The historical query module is used to store and retrieve historical data and statistical analysis results.

8. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-6.

9. A computer storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-6.

Images