Interface layout generation method based on intelligent gateway and intelligent gateway

Abstract

The invention relates to an interface layout generation method based on an intelligent gateway and the intelligent gateway, and relates to the technical field of Internet of Things, and the interface layout generation method comprises the steps: obtaining and transmitting an instrument monitoring video to the intelligent gateway, and recognizing an instrument reading; according to an equipment fault dynamic judgment mechanism, combining with sensor data, fusing and analyzing instrument reading to obtain a data analysis result, and activating a corresponding operation trigger point; determining an interface function weight according to the operation trigger point in combination with a preset weight evaluation model, and constructing an interface layout template in combination with a dynamic layout generation algorithm; capturing and coding user operation behaviors, generating an interface operation event, and correcting the interface layout template in combination with event processing logic to obtain an optimal interface layout; an interface function weight is determined by constructing a weight evaluation model, a reasonable same-screen layout is automatically generated according to different business requirements and use scenes in combination with a dynamic layout generation algorithm and an interface layout template storage calling mechanism, and efficient and visual information display is obtained.

Description

Technical Field

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

[0002] In fields such as power generation and industrial automation, gateway devices serve as key nodes for data transmission and processing, evolving from simple data forwarding to intelligent analysis and processing. Traditional gateways primarily focus on data transmission efficiency and stability. While they possess basic data transmission capabilities, they still suffer from single-function, low hardware integration, and inconsistent communication protocols. They also lack intelligent data analysis, AI recognition, user-friendly interfaces, and protocol compatibility. They also lack integrated intelligent analysis modules, such as pointer and digital meter recognition, as well as intuitive user-friendly human-computer interaction interfaces.

[0003] Patent No. CN202410750395.0 discloses a method, system and storage medium for automatically generating an interface layout. The method for automatically generating an interface layout is loaded into an existing in-vehicle intelligent system to directly obtain the interface resolution parameters of the current interface and calculate the style configuration parameters. Then, the layout parameters of the basic grid in the current interface layout are calculated in combination with the element parameters corresponding to the current interface style, thereby determining the layout parameters of the current display interface. At this time, the interface proportion of each application interface is allocated based on the current scene and / or user characteristics, and the basic grid is partitioned to obtain interface partitions, and the application interface is laid out according to user needs. Subsequently, the layout coordinates of each element in the application interface are further calculated, thereby automatically generating the layout and elements of the UI screen according to user preferences, changes in usage scenarios, etc., giving users a flexible and diverse intelligent experience that suits their needs.

[0004] The above-mentioned prior art solutions have the following defects: 1. Traditional gateways lack intelligent analysis capabilities and cannot effectively identify pointer instrument and digital instrument information, reducing the level of intelligent data processing.

[0005] 2. The existing technology has a high failure rate and poor scalability due to the dispersion of hardware interfaces, and lacks efficient and stable protocol support for communication with the upper-level master station. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention determines the interface function weights by constructing a weight evaluation model, and combines the dynamic layout generation algorithm and the interface layout template storage and call mechanism to automatically generate a reasonable same-screen layout according to different business needs and usage scenarios, thereby obtaining efficient and intuitive information display.

[0007] This is achieved using the following technical solutions: A method for generating an interface layout based on an intelligent gateway, comprising: Acquire and transmit instrument monitoring videos to the intelligent gateway and identify instrument readings; According to the dynamic judgment mechanism of equipment failure, combined with sensor data, the instrument readings are integrated and analyzed to obtain data analysis results and activate corresponding operation trigger points; Determine the interface function weights based on the operation trigger points in combination with a preset weight evaluation model, and construct an interface layout template in combination with a dynamic layout generation algorithm; Capture and encode user operation behaviors, generate interface operation events, and modify the interface layout template in combination with event processing logic to obtain the optimal interface layout.

[0008] By adopting the above technical solution, the intelligent gateway and SIP-B protocol transmission technology are integrated to obtain instrument monitoring video, and filtering enhancement, edge detection and OCR algorithms are used to realize pointer angle calculation and digital character recognition. A multi-source feature fusion model is constructed in combination with sensor data, and machine learning is used to realize dynamic fault prediction and trigger operation instructions. The interface business priority is quantified based on the weight evaluation model, and the visual focus area is automatically allocated through the dynamic layout generation algorithm combined with screen parameters. A layout template library is established to realize scene adaptive switching. At the same time, the event monitoring mechanism is used to capture user behavior characteristics, and the operation response link is optimized through coding mapping and filtering rules. Combined with the animation feedback mechanism and hierarchical data push strategy, a "perception-analysis-decision-making-interaction" closed loop is formed to achieve the three goals of accurate instrument reading identification, real-time fault warning, and dynamic optimization of the operation interface. It has the characteristics of high intelligence level, strong multi-protocol compatibility, and excellent interface adaptation efficiency.

[0009] The present invention is further configured such that the specific steps of acquiring and transmitting the instrument monitoring video to the intelligent gateway and identifying the instrument readings include: Monitor all instruments and obtain initial instrument monitoring videos; Dividing the initial instrument monitoring video into blocks to obtain a plurality of instrument video blocks; Perform authority authentication between the gateway host and the main station receiving end, complete the communication connection, and build a transmission channel; Encrypting and encapsulating all the instrument video blocks according to a preset redundant link protocol to obtain a plurality of encrypted and encapsulated video blocks; transmitting the encrypted and encapsulated video blocks to the master station receiving end through the transmission channel, and monitoring the data transmission process to obtain a network status and a transmission rate; Modifying the transmission rate according to the network status and verifying the integrity of each encrypted and encapsulated video block in combination with a cache retransmission mechanism; If all the encrypted and encapsulated video blocks are complete and correct, decrypting the encrypted and encapsulated video blocks and splicing them according to the timestamps to obtain the aggregated instrument video; Matching and frame decomposing the aggregated instrument video according to a preset instrument template to obtain an instrument time sequence frame; Sampling the instrument time sequence frame according to a preset time interval to obtain an original instrument image; Filtering and denoising the original instrument image to obtain a noise-free instrument image; performing image enhancement on the noise-free instrument image to obtain an enhanced instrument image; Classifying the enhanced instrument image according to instrument type to obtain a pointer instrument image and a digital instrument image; Performing edge detection on the pointer instrument image to obtain a dial outline and pointer lines; Fitting the dial outline and the pointer line according to a preset angle calculation model and the instrument reference scale to identify the instrument reading; Performing affine transformation and edge detection on the digital instrument image to obtain a character display frame; Horizontally positioning and vertically segmenting the character display frame according to pixel continuity to obtain a plurality of independent characters; Each independent character is identified and combined according to the preset character recognition model to obtain the instrument reading.

[0010] By adopting the above technical solution, instrument videos are collected in real time through monitoring cameras and encrypted in blocks for transmission to the intelligent gateway. A transmission channel is built based on the redundant link protocol, and integrity verification and dynamic rate adjustment are performed. The aggregated video is classified into instrument types using the YOLOv7 model. Edge detection and angle fitting algorithms are used to identify pointer instrument readings, and digital instrument data is parsed through affine transformation and character segmentation models. This realizes the automatic and accurate collection and transmission of instrument data in industrial scenarios. It has the advantages of efficient transmission, multi-protocol compatibility, dynamic anti-interference and abnormal alarm, which can effectively reduce the cost of manual inspections and improve the efficiency of equipment status monitoring.

[0011] The present invention is further configured as follows: the specific steps of combining the instrument readings with the sensor data according to the dynamic judgment mechanism of equipment failure to obtain the data analysis results and activate the corresponding operation trigger point include: The running time of the equipment in different working modes is counted and arranged in descending order to obtain the working time sequence of the working mode; Correlate and analyze the working condition time series with historical sensor data according to the timestamp to obtain working condition correlation time series data; determine a static judgment threshold interval based on the equipment characteristics and the working condition correlation time series data; Perform fitting calculation on the static judgment threshold intervals of different working modes to obtain the dynamic function of all working conditions; Perform feature extraction on instrument readings and sensor data to obtain instrument characteristic values ​​and sensor characteristic values; Comparing the instrument characteristic value and the sensor characteristic value with the static judgment threshold interval respectively; if either one is not within the static judgment threshold interval, determining that the device is suspected of failure and generating a corresponding failure decision value; if both are not within the static judgment threshold interval, determining that the device has failed; Generate a dynamic determination threshold interval based on the operating mode and the full operating condition dynamic function, and compare and determine with the instrument characteristic value and the sensor characteristic value respectively; if any one of them is not within the dynamic determination threshold interval, modify the corresponding fault decision weight coefficient; if both are not within the dynamic determination threshold interval, perform a level determination on the equipment fault and generate a fault level code; Generate a fault trigger value based on the fault decision weight coefficient and the fault decision value, and match it with a preset fault level warning interval to obtain the fault level code; According to the fault level code and the fault duration, the corresponding operation trigger point is activated.

[0012] By adopting the above technical solution, a fault diagnosis algorithm based on time series aggregation analysis and dynamic threshold fitting first constructs the working time series of the working mode by statistically arranging the operating time in descending order. This is then associated with historical sensor data to generate working condition-related time series data. Static threshold intervals are extracted based on equipment characteristics and a dynamic function for all working conditions is fitted. Faults are then determined by dual comparison of instrument and sensor characteristic values ​​with static / dynamic thresholds (static thresholds trigger preliminary fault decision values, while dynamic thresholds correct decision weights). Finally, the weight coefficients are integrated with the fault level warning interval to generate a fault level code and activate the corresponding action trigger point. Its advantage lies in the integration of time series clustering, multi-stage statistical modeling, static time series anomaly detection, and dynamic function fitting technologies. Through a two-tiered judgment mechanism of static threshold initial screening and dynamic threshold adaptive calibration, it significantly improves fault identification accuracy and reduces false alarm rates in multiple working condition scenarios. At the same time, relying on time series data association and weight correction strategies, the system's adaptability to nonlinear working condition changes is enhanced, achieving intelligent closed-loop control of fault classification response and resource scheduling.

[0013] The present invention is further configured as follows: the specific steps of determining the interface function weights based on the operation trigger points in combination with a preset weight evaluation model, and constructing the interface layout template in combination with a dynamic layout generation algorithm include: Based on user operation frequency and data update frequency combined with heat maps, the operation interface is divided and high-frequency touch areas are marked; high-frequency touch areas are disassembled according to business processes, key trigger actions are marked, and functional interface areas are divided; According to the business value and information priority, indicators are constructed for the functional interface area to obtain a functional trigger matrix; Responsive rules are set according to the interface size, and the function trigger matrix is ​​iteratively trained in combination with the operation trigger point to obtain a weight evaluation model and output the interface function weight; Allocate all functional areas according to the interface function weights in combination with a dynamic layout generation algorithm to obtain corresponding interface space proportions; According to the interface space proportion and Gestalt principles, all the functional areas are clustered and deployed to obtain an interface layout template.

[0014] By adopting the above technical solution, the high-frequency touch areas are divided through the K-means clustering algorithm of the heat map, the function trigger matrix is ​​established in combination with business process decomposition, the hierarchical analysis method is used to construct business value indicators, the BP neural network is iteratively trained under responsive rules to generate a weight evaluation model, the interface space ratio is allocated based on the dynamic layout genetic algorithm, and finally the interface layout template is realized through Gestalt perception clustering to form a data-driven adaptive interface optimization system. Its advantage lies in dynamically adjusting the layout weight through machine learning, combining user behavior analysis with cognitive psychology principles to achieve spatial matching between interface functional areas and operating habits, while ensuring business process efficiency, significantly improving visual hierarchy perception, improving the efficiency of high-frequency function triggering, and having cross-terminal responsiveness.

[0015] 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 the operation trigger point to obtain a weight evaluation model. The specific steps of outputting the interface function weight include: Divide the operation interface into preset size groups according to the interface size, define the number of grid columns and component priority; Generate responsive rules based on the number of grid columns and the component priority combined with layout adaptation logic; According to the responsive rules and the operation trigger points, weights are assigned to the function trigger matrix to form a weight vector matrix; According to the preset constraint parameters and the weight vector matrix, the function trigger matrix is ​​iteratively trained several times to obtain a weight evaluation model; Optimizing the initial weight matrix W of the weight evaluation model according to the target loss function L to obtain the optimal weight matrix; Among them, 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 functional module, s j is the device size weight coefficient, M[i,j] is the function trigger matrix element, λ is the regularization coefficient, and Reg is the regularization term; The operation interface is weighted according to the optimal weight matrix combined with user behavior data, and the interface function weight is output.

[0016] By adopting the above technical solution, the interface is divided based on preset size groups and the grid dynamic programming algorithm is used to define the number of columns and component priorities. The function trigger matrix is ​​constructed by combining responsive breakpoint rules and operation heat map data. By introducing the device size weight coefficient and the L2 regularization optimization objective function with regularization term constraints, the weight evaluation model is iteratively trained using the gradient descent algorithm. Finally, the layout is dynamically adjusted using a genetic algorithm to generate the optimal weight matrix, realizing the weight distribution of multi-terminal adaptive interface functions. Its advantage lies in avoiding overfitting by controlling the complexity of the model through regularization, dynamically optimizing the space share by combining device characteristics and user behavior, ensuring the matching of high-frequency function layout and visual movement lines in cross-resolution scenarios, improving interface operation efficiency, and reducing multi-terminal adaptation costs based on grid elastic scaling.

[0017] The present invention is further configured as follows: the specific steps of capturing and encoding user operation behaviors, generating interface operation events, and modifying the interface layout template in combination with event processing logic to obtain the optimal interface layout include: Capture and count user operation behaviors on the operation interface to obtain interface operation events; Filter all the interface operation events for invalid operations according to the event monitoring mechanism to obtain valid operation events; Classifying and coding the valid operation events to obtain unique event codes; Sort the functional areas according to the unique event code and event processing logic to generate a function triggering process; The function triggering process is disassembled and associated with the functional modules of the device to obtain a process mapping relationship table; according to the process mapping relationship table and the user's operating habits, the interface layout template is modified to obtain the optimal interface layout.

[0018] By adopting the above technical solution, user operation behaviors are captured through the event monitoring mechanism and invalid operations are filtered using a dynamic threshold algorithm. Valid operation events are classified and encoded using a clustering algorithm to generate a unique identifier. The hidden Markov model (HMM) is used to build a function trigger process and a process mapping relationship table is built based on association rule mining. Finally, a genetic algorithm or a reinforcement learning model is used to dynamically optimize the interface layout to achieve adaptive interface sorting adjustments based on user operation habits. The advantage is that through a data-driven closed-loop optimization mechanism, it reduces the user's cognitive load while improving operational efficiency, and has real-time response and personalized adaptation capabilities.

[0019] In a second aspect, the present invention further provides an interface layout generation system based on an intelligent gateway, which adopts the following technical solutions: An interface layout generation system based on an intelligent gateway, comprising: The sensor monitoring module is used to display the device deployment environment parameter data collected by sensors in real time; AI recognition module, used to analyze and identify pointer instruments 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 equipment; Status monitoring module, used to monitor the operating status and abnormal alarm information of the smart gateway; Video surveillance module, used to display the station house video stream and remote instrument monitoring screen in real time; Intelligent inspection module, used to automatically perform inspection tasks and provide feedback on equipment operating status; AI analysis module, used to integrate multi-source data to achieve fault prediction and intelligent decision-making analysis; Linkage control module, used to trigger equipment control and emergency response operations according to preset logic; The historical query module is used to store and review historical data and statistical analysis results.

[0020] By adopting the above technical solution, the sensor monitoring module collects environmental parameters (such as temperature, humidity, and light) in real time. The AI ​​recognition module uses the convolutional neural network (CNN) algorithm to analyze instrument images and obtain accurate readings. The topology optimization algorithm of the floor plan module is combined with the visualization of equipment location and connection relationships. The time series anomaly detection algorithm of the condition monitoring module diagnoses the gateway operation status in real time and triggers alarms. The multi-target tracking algorithm of the integrated video surveillance module simultaneously presents the real-time image of the station building. The path planning algorithm of the intelligent inspection module autonomously executes inspection tasks. The machine learning models (such as LSTM and random forest) of the AI ​​analysis module are integrated to achieve multi-source data-driven fault prediction and decision optimization. The rule engine of the linkage control module dynamically adjusts equipment operation strategies. The time series database and visualization algorithm of the historical query module are combined to trace data trends. Ultimately, an intelligent management system integrating monitoring, identification, analysis, and control is established. This significantly improves management efficiency and decision-making accuracy, reduces the need for manual intervention, enhances system real-time performance and reliability, and provides full lifecycle support for equipment operation and maintenance. At the same time, resource allocation is optimized through multimodal data fusion and adaptive logic, improving fault warning accuracy and inspection efficiency.

[0021] In a third aspect, the present invention further provides an electronic device, which adopts the following technical solution: An electronic device, comprising: one or more processors; a memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the above method.

[0022] By adopting the above technical solution, the above-mentioned interface layout generation method based on the intelligent 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 above-mentioned interface layout generation method based on the intelligent gateway are executed to achieve the effect of reducing manual labor intensity and improving the degree of automated operation.

[0023] In a fourth aspect, the present invention further provides a computer storage medium, which adopts the following technical solution: A computer storage medium stores a computer program, which implements the above method when executed by a processor.

[0024] In summary, the beneficial technical effects of the present invention are: 1. By building a weight evaluation model to determine the weight of interface functions, and combining it with a dynamic layout generation algorithm and a layout template storage and call mechanism, it can automatically generate a reasonable same-screen layout based on different business needs and usage scenarios, achieving efficient and intuitive information display.

[0025] 2. Classify and encode operation events, establish a mapping relationship with functional modules, and cooperate 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.

[0026] 3. Based on data update trigger monitoring, the update information is prioritized and a hierarchical push and display strategy is adopted to enable the device to intelligently process information of different importance levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flowchart of an interface layout generation method according to one embodiment of the present invention.

[0028] Figure 2 It is a flowchart of an interface layout generation method according to one embodiment of the present invention.

[0029] Figure 3 It is a flowchart of an interface layout generation method according to one embodiment of the present invention.

[0030] Figure 4 It is a structural diagram of an interface layout generation system according to one embodiment of the present invention. DETAILED DESCRIPTION

[0031] The present invention will be further described in detail below with reference to the accompanying drawings.

[0032] Reference Figure 1 , is a method for generating an interface layout based on an intelligent gateway disclosed in the present invention, comprising: S1: Acquire and transmit instrument monitoring video to the smart gateway and identify instrument readings; S2: Based on the dynamic judgment mechanism of equipment failure and sensor data, the instrument readings are integrated and analyzed to obtain data analysis results and activate corresponding operation trigger points; S3: Determine the interface function weights based on the operation trigger points and a preset weight evaluation model, and construct an interface layout template based on a dynamic layout generation algorithm; S4: Capture and encode user operation behaviors, generate interface operation events, and modify the interface layout template in combination with event processing logic to obtain the optimal interface layout.

[0033] The implementation principles of this embodiment are as follows: receiving multiple instrument monitoring video streams through an intelligent gateway, using a dynamic bitrate adjustment algorithm to ensure transmission stability, and identifying readings based on an improved YOLOv7 visual model; synchronously accessing multi-band sensor data, realizing multi-source information fusion through timestamp alignment and Kalman filtering, and combining the fault knowledge graph with the Bayesian network to generate dynamic operation trigger points; the weight evaluation model analyzes user historical behavior data through a reinforcement learning module, dynamically optimizes function priorities, and drives a genetic algorithm to generate the initial interface layout; user operation behavior is encoded in real time by the edge node, triggering an event-driven engine, which iteratively modifies layout parameters based on operation success rate, latency, and preference tags, and ultimately outputs the optimal interface adapted to the terminal through a distributed rendering service.

[0034] Example 2: The specific steps of step S1 include: Monitor all instruments and obtain initial instrument monitoring videos; Dividing the initial instrument monitoring video into blocks to obtain a plurality of instrument video blocks; Perform authority authentication between the gateway host and the main station receiving end, complete the communication connection, and build a transmission channel; Encrypting and encapsulating all the instrument video blocks according to a preset redundant link protocol to obtain a plurality of encrypted and encapsulated video blocks; transmitting the encrypted and encapsulated video blocks to the master station receiving end through the transmission channel, and monitoring the data transmission process to obtain a network status and a transmission rate; Modifying the transmission rate according to the network status and verifying the integrity of each encrypted and encapsulated video block in combination with a cache retransmission mechanism; In this embodiment, the video information is encapsulated according to the SIP-B protocol at the gateway host end, and the main station receiving end performs parsing and processing; a handshake protocol is used to establish a connection and authenticate permissions, the connection status is monitored in real time, and automatic reconnection is used to ensure continuous transmission; the transmission rate and encoding format are dynamically adjusted according to the network conditions, and the cache and retransmission mechanism is used to ensure data integrity and accuracy.

[0035] In order to solve the problem of the gateway's compatibility with multiple communication protocols, especially the insufficient support for the SIP-B protocol, a highly modular design concept is adopted to integrate multiple interface resources, including POE network, storage, serial port, low current sampling, input and output signal control, 4G / 5G communication, LORA, WIFI, Bluetooth, etc.

[0036] If all the encrypted and encapsulated video blocks are complete and correct, decrypting the encrypted and encapsulated video blocks and splicing them according to the timestamps to obtain the aggregated instrument video; Matching and frame decomposing the aggregated instrument video according to a preset instrument template to obtain an instrument time sequence frame; Sampling the instrument time sequence frame according to a preset time interval to obtain an original instrument image; Filtering and denoising the original instrument image to obtain a noise-free instrument image; performing image enhancement on the noise-free instrument image to obtain an enhanced instrument image; Classifying the enhanced instrument image according to instrument type to obtain a pointer instrument image and a digital instrument image; Performing edge detection on the pointer instrument image to obtain a dial outline and pointer lines; Fitting the dial outline and the pointer line according to a preset angle calculation model and the instrument reference scale to identify the instrument reading; Performing affine transformation and edge detection on the digital instrument image to obtain a character display frame; Horizontally positioning and vertically segmenting the character display frame according to pixel continuity to obtain a plurality of independent characters; Each independent character is identified and combined according to the preset character recognition model to obtain the instrument reading.

[0037] In this embodiment, the built-in image acquisition module is used to collect instrument images, which are then processed through filtering and enhancement algorithms to improve image quality. The dial outline and pointer lines are extracted through edge detection, and the pointer value is determined by combining the angle calculation model, and the scale is calibrated. OCR technology is used to segment characters, extract features, and match patterns to establish and optimize a character template library.

[0038] The implementation principles of this embodiment are as follows: the timestamps of multi-channel instrument videos and sensor data are aligned through the SIP-B protocol, the SRT protocol is superimposed with quantum keys to perform dual-link encrypted transmission of block video streams, and the image distortion caused by mechanical vibration is dynamically compensated in combination with MEMS inertial sensors; during the transmission process, data integrity is verified based on the blockchain hash chain, and in the event of anomalies, the local cache of the edge node and TEE encrypted retransmission are triggered; at the receiving end, the improved CRNN model is driven to recognize pointer / digital instrument readings through multi-spectral fusion enhancement and the anti-interference character library generated by GAN; at the same time, the video bit rate and encryption strength are dynamically adjusted according to the network status, and finally, the aggregated video and high-confidence reading data that have undergone spatiotemporal consistency verification are synchronously injected into the knowledge graph.

[0039] Example 3: The specific steps of step S2 include: The running time of the equipment in different working modes is counted and arranged in descending order to obtain the working time sequence of the working mode; The operating condition time series is correlated with historical sensor data based on the timestamps to generate the associated operating condition time series data. In this embodiment, the priority of the primary and backup sensors can also be set, and the initial operating mode and data collection frequency can be defined. During the initialization phase, the normal operating range, error tolerance, and fault determination criteria (such as data fluctuation thresholds and abnormal duration) of each sensor must be clearly defined.

[0040] Determine a static judgment threshold interval based on the equipment characteristics and the time series data associated with the working condition; Perform fitting calculation on the static judgment threshold intervals of different working modes to obtain the dynamic function of all working conditions; In this embodiment, static thresholds (such as ±5°C deviation of a temperature sensor) or dynamic thresholds (such as sliding window statistics based on a time series) are preset based on historical data and equipment characteristics and calibrated to adapt to different working conditions.

[0041] Perform feature extraction on instrument readings and sensor data to obtain instrument characteristic values ​​and sensor characteristic values; Comparing the instrument characteristic value and the sensor characteristic value with the static judgment threshold interval respectively; if either one is not within the static judgment threshold interval, determining that the device is suspected of failure and generating a corresponding failure decision value; if both are not within the static judgment threshold interval, determining that the device has failed; In this embodiment, the output data of each sensor is monitored in real time and compared with a preset threshold range. If the data exceeds the threshold continuously (for example, exceeding the limit within three sampling periods), a primary fault alarm is triggered and the sensor is marked as "suspicious state".

[0042] Generate a dynamic determination threshold interval based on the operating mode and the full operating condition dynamic function, and compare and determine with the instrument characteristic value and the sensor characteristic value respectively; if any one of them is not within the dynamic determination threshold interval, modify the corresponding fault decision weight coefficient; if both are not within the dynamic determination threshold interval, perform a level determination on the equipment fault and generate a fault level code; In this embodiment, the operating modes include normal operation (parameters are stable within the design range and the equipment works efficiently), attention status (parameters are approaching the limit but not exceeding the limit, and monitoring needs to be strengthened), abnormal status (important parameters are close to or slightly exceed the limit, and maintenance needs to be arranged) and serious status (key parameters are seriously exceeded and the equipment must be shut down for maintenance immediately).

[0043] In this embodiment, redundant sensor data can also be cross-validated to identify spatial consistency of abnormal data. For example, if the primary sensor temperature value is abnormal but the backup sensor data is normal, the primary sensor is considered faulty. If both are abnormal, it may be due to a real change in the environment or a system-level failure.

[0044] Generate a fault trigger value based on the fault decision weight coefficient and the fault decision value, and match it with a preset fault level warning interval to obtain the fault level code; In this embodiment, fault levels are also divided according to the degree and duration of deviation from the threshold (such as slight drift and severe failure). For example: Level 1 fault: Data fluctuations are within the 10% threshold, triggering an alert but not switching devices. Level 2 fault: If the threshold is exceeded by 20% and persists for 5 seconds, a master / slave switchover is triggered. Level 3 fault: Multiple sensors are abnormal at the same time, triggering an emergency shutdown of the system.

[0045] Automatically switch to the backup sensor based on the preset priority, update the control rights, and feedback the switch status to the host computer. The switching process must ensure the continuity of the control signal (such as using a smooth transition algorithm) to avoid system oscillation.

[0046] According to the fault level code and the fault duration, the corresponding operation trigger point is activated.

[0047] The implementation principle of this embodiment is as follows: the controller captures the equipment electrical signal jump points in real time, combines with the edge computing node to clean the time series data and record the start and stop time of the working condition; based on the LSTM model, historical sensor data is analyzed to generate a dynamic threshold function, and the Kalman filter is simultaneously used to fuse the main and standby sensor data to eliminate local anomalies; when the instrument characteristic value deviates from the threshold, multi-spectral verification and spatial consistency analysis are used to distinguish sensor failure or equipment abnormality, and the MES work order distribution and visual large-screen alarm are triggered according to the fault level code; the main and standby switching process is embedded with a PID smoothing algorithm to ensure control continuity, and at the same time, the fault decision records are stored based on the blockchain to realize maintenance traceability. Finally, the threshold interval is iteratively optimized through the adaptive learning engine to form an intelligent monitoring system with equipment status perception-dynamic threshold judgment-multi-level fault response-closed-loop knowledge update.

[0048] Example 4: Reference Figure 2 , the specific steps of step S3 include: S31: Divide the operation interface according to the user operation frequency and data update frequency in combination with the heat map, and mark the high-frequency touch area; S32: Disassemble high-frequency touch areas according to business processes, mark key trigger actions, and divide functional interface areas; In this embodiment, key trigger actions include button clicks and form submissions; the functional interface areas include the primary functional area (such as payment), the secondary functional area (such as help entry), and the extended reserved area (such as sharing); S33: constructing indicators for the functional interface area based on business value and information priority to obtain a function trigger matrix; S34: setting responsive rules based on interface size, and iteratively training the function trigger matrix in combination with operation trigger points to obtain a weight evaluation model and output interface function weights; S35: Allocating all functional areas according to the interface function weights in combination with a dynamic layout generation algorithm to obtain corresponding interface space proportions; In this embodiment, the interface space ratios are as follows: the main function area occupies 40%-60% of the screen, using a fixed grid layout (such as the 12-column Bootstrap system), the secondary function area occupies 20%-30%, using a flexible layout (Flexbox) to adapt to dynamic content, and the extended reserved area retains a 10%-15% space margin to support subsequent function iterations.

[0049] S36: Clustering and deploying all the functional areas according to the interface space proportion and Gestalt principles to obtain an interface layout template.

[0050] The implementation principle of this embodiment is as follows: by integrating multi-dimensional user behavior data (including spatiotemporal distribution heat maps and scenario-based operation flows) with interactive physical rules, a dynamic function trigger matrix is ​​constructed. Combined with responsive device adaptation and personalized profiling, an initial layout is generated using a Gestalt visual clustering algorithm. Subsequently, based on the real-time operational data backflow, the function weights are iteratively optimized through an online learning model, driving the elastic layout engine to allocate space according to the proportion of primary and secondary functional areas. A / B testing modules are simultaneously embedded to verify visual movement and operational efficiency, ultimately forming an adaptive interface system that balances business goals, user experience, and iterative scalability. The key is to combine statistical laws, cognitive psychology principles, and dynamic machine learning parameter adjustment to achieve a closed-loop optimization of "data perception-rule reasoning-layout generation-effect verification," ensuring that interface space allocation always matches high-frequency user needs and the direction of business evolution.

[0051] Embodiment 5: Reference Figure 3 The specific steps of step S34 include: S341: Divide the operation interface into preset size groups according to the interface size, and define the number of grid columns and component priority; In this embodiment, the preset size groups include mobile <768px, tablet 768-1024px, and desktop >1024px. Each group defines layout rules (such as the number of grid columns and component priority); S342: Generate responsive rules based on the number of grid columns and the component priority in combination with layout adaptation logic; S343: assigning weights to the function trigger matrix according to the responsive rule in combination with the operation trigger point to form a weight vector matrix; S344: performing several iterative training on the function trigger matrix according to preset constraint parameters in combination with the weight vector matrix to obtain a weight evaluation model; S345: Optimizing the initial weight matrix W of the weight evaluation model according to the target loss function L to obtain an optimal weight matrix; Among them, 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 functional module, s j is the device size weight coefficient, M[i,j] is the function trigger matrix element, λ is the regularization coefficient, and Reg is the regularization term; S346: Allocate weights to the operation interface based on the optimal weight matrix and user behavior data, and output interface function weights.

[0052] The implementation principle of this embodiment is as follows: a dynamic function trigger matrix is ​​constructed by integrating multi-dimensional user behavior data (including spatiotemporal distribution heat maps and device input modal characteristics) with interactive physical laws, differentiated responsive rules are generated based on the device recognition module and the scene state machine, and personalized weight distribution is achieved by combining the attenuation factor; the target loss function embedded in the visual inertia term is used to drive the iteration of the weight evaluation model, and the real-time operation data is absorbed through the online learning mechanism to optimize the matrix parameters. Finally, relying on the elastic layout engine with Gestalt continuity constraints, an adaptive interface that meets business goals, interaction efficiency and cognitive consistency is output.

[0053] Example 6: The specific steps of step S4 include: Capture and count user operation behaviors on the operation interface to obtain interface operation events; Filter all the interface operation events for invalid operations according to the event monitoring mechanism to obtain valid operation events; Classifying and coding the valid operation events to obtain unique event codes; In this embodiment, operation events are classified and encoded, a monitoring mechanism is used to capture and filter invalid events in real time, functional modules are called for processing through mapping relationships, operation feedback is provided through animation and sound, trigger points are set, update information is prioritized, a hierarchical push and display strategy is adopted, and user-defined settings are supported.

[0054] Sort the functional areas according to the unique event code and event processing logic to generate a function triggering process; The function triggering process is disassembled and associated with the functional modules of the device to obtain a process mapping relationship table; according to the process mapping relationship table and the user's operating habits, the interface layout template is modified to obtain the optimal interface layout.

[0055] The implementation principle of this embodiment is as follows: a dynamic threshold algorithm based on time windows and trajectory curvature is used, and system-level permissions are integrated to realize global touch event monitoring; effective operations are captured through global touch event monitoring and dynamic threshold filtering, and high-concurrency processing is achieved based on standardized coding rules and asynchronous event queues; the device input modality recognition layer is combined to adapt to multi-terminal interaction differences, and the sliding time window and LSTM model are used to dynamically update the user habit weights to drive functional process mapping optimization; interaction fault tolerance is enhanced through composite feedback channels and transactional logs, and finally, relying on thread priority scheduling and breakpoint resumption strategies, an adaptive interface that takes into account real-time response, cross-terminal compatibility and behavior prediction accuracy is output, thereby realizing dynamic optimal matching of user intentions and interface responses.

[0056] Embodiment seven: Reference Figure 4 , an interface layout generation system based on intelligent gateway, comprising: The sensor monitoring module is used to display the device deployment environment parameter data collected by sensors in real time; AI recognition module, used to analyze and identify pointer instruments 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 equipment; Status monitoring module, used to monitor the operating status and abnormal alarm information of the smart gateway; Video surveillance module, used to display the station house video stream and remote instrument monitoring screen in real time; Intelligent inspection module, used to automatically perform inspection tasks and provide feedback on equipment operating status; AI analysis module, used to integrate multi-source data to achieve fault prediction and intelligent decision-making analysis; Linkage control module, used to trigger equipment control and emergency response operations according to preset logic; The historical query module is used to store and review historical data and statistical analysis results.

[0057] The implementation principle of this embodiment is: first, based on the business importance, data update frequency and operation usage frequency of nine modules such as sensor monitoring and AI recognition, a weight evaluation model is constructed to determine the interface function weight. According to the interface function weight, a dynamic layout generation algorithm is designed. High-weight interfaces are allocated in the center of the screen or in the visual focus area, and low-weight interfaces are placed on the edge or folded, hidden, etc. A layout template library is established to store layout schemes generated under different weight combinations and screen parameters. When the system starts or the user switches the usage scenario, the corresponding layout template is quickly called according to the current interface weight and device parameters to achieve efficient initialization and switching of the operation interface.

[0058] Classify user actions on the interface, such as clicks, swipes, long presses, and pinch-to-zoom, and assign unique event codes. Create a mapping table between action events and functional modules, clearly defining the system functions and processing logic corresponding to each action event. Leverage the operating system's event monitoring mechanism to capture user actions in real time. Set event filtering rules to filter out duplicate and invalid action events to avoid wasting system resources. For example, if the same area is clicked multiple times within a short period of time, only the first valid action will be retained. When the system receives a valid action event, it calls the corresponding functional module for processing based on the event code and mapping table.

[0059] Data update trigger points are set in the data acquisition module and system status monitoring module of the gateway device. When the sensor collects new data, the instrument identification result changes, or the system status becomes abnormal, a data update event is triggered, and the type, source, and other information of the updated data are encapsulated. Update information is prioritized based on the importance and urgency of the data. For high-priority information, strong reminders such as full-screen pop-ups and high-frequency flashing are used to push it, and it is displayed in a prominent position on the operation interface. For medium and low priority information, it is pushed through lightweight methods such as notification bar prompts and icon badges, and the data is updated on the corresponding interface.

[0060] An electronic device, comprising: one or more processors; a memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the above method.

[0061] A computer storage medium stores a computer program, which implements the above method when executed by a processor.

[0062] The embodiments of this specific implementation method are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for generating an interface layout based on an intelligent gateway, characterized in that: include: Acquire and transmit instrument monitoring videos to the intelligent gateway and identify instrument readings; According to the dynamic judgment mechanism of equipment failure, combined with sensor data, the instrument readings are integrated and analyzed to obtain data analysis results and activate corresponding operation trigger points; Determine the interface function weights based on the operation trigger points in combination with a preset weight evaluation model, and construct an interface layout template in combination with a dynamic layout generation algorithm; Capture and encode user operation behaviors, generate interface operation events, and modify the interface layout template in combination with event processing logic to obtain the optimal interface layout.

2. The method for generating an interface layout based on an intelligent gateway according to claim 1, characterized in that: The specific steps of acquiring and transmitting the instrument monitoring video to the intelligent gateway and identifying the instrument readings include: Monitor all instruments and obtain initial instrument monitoring videos; Dividing the initial instrument monitoring video into blocks to obtain a plurality of instrument video blocks; Perform authority authentication between the gateway host and the main station receiving end, complete the communication connection, and build a transmission channel; Encrypting and encapsulating all the instrument video blocks according to a preset redundant link protocol to obtain a plurality of encrypted and encapsulated video blocks; transmitting the encrypted and encapsulated video blocks to the master station receiving end through the transmission channel, and monitoring the data transmission process to obtain a network status and a transmission rate; Modifying the transmission rate according to the network status and verifying the integrity of each encrypted and encapsulated video block in combination with a cache retransmission mechanism; If all the encrypted and encapsulated video blocks are complete and correct, the encrypted and encapsulated video blocks are decrypted and spliced ​​together according to the timestamps to obtain the aggregated instrument video.

3. The method for generating an interface layout based on an intelligent gateway according to claim 2, characterized in that: The specific steps of acquiring and transmitting the instrument monitoring video to the intelligent gateway and identifying the instrument readings also include: Matching and frame decomposing the aggregated instrument video according to a preset instrument template to obtain an instrument time sequence frame; Sampling the instrument time sequence frame according to a preset time interval to obtain an original instrument image; Filtering and denoising the original instrument image to obtain a noise-free instrument image; performing image enhancement on the noise-free instrument image to obtain an enhanced instrument image; Classifying the enhanced instrument image according to instrument type to obtain a pointer instrument image and a digital instrument image; Performing edge detection on the pointer instrument image to obtain a dial outline and pointer lines; Fitting the dial outline and the pointer line according to a preset angle calculation model and the instrument reference scale to identify the instrument reading; Performing affine transformation and edge detection on the digital instrument image to obtain a character display frame; Horizontally positioning and vertically segmenting the character display frame according to pixel continuity to obtain a plurality of independent characters; Each independent character is identified and combined according to the preset character recognition model to obtain the instrument reading.

4. The method for generating an interface layout based on an intelligent gateway according to claim 1, characterized in that: The specific steps of combining the dynamic judgment mechanism of equipment failure with sensor data, fusing and analyzing the instrument readings, obtaining data analysis results, and activating corresponding operation trigger points include: The running time of the equipment in different working modes is counted and arranged in descending order to obtain the working time sequence of the working mode; Correlate and analyze the working condition time series with historical sensor data according to the timestamp to obtain working condition correlation time series data; determine a static judgment threshold interval based on the equipment characteristics and the working condition correlation time series data; Perform fitting calculation on the static judgment threshold intervals of different working modes to obtain the dynamic function of all working conditions; Perform feature extraction on instrument readings and sensor data to obtain instrument characteristic values ​​and sensor characteristic values; Comparing the instrument characteristic value and the sensor characteristic value with the static judgment threshold interval respectively; if either one is not within the static judgment threshold interval, determining that the device is suspected of failure and generating a corresponding failure decision value; if both are not within the static judgment threshold interval, determining that the device has failed; Generate a dynamic determination threshold interval based on the operating mode and the full operating condition dynamic function, and compare and determine with the instrument characteristic value and the sensor characteristic value respectively; if any one of them is not within the dynamic determination threshold interval, modify the corresponding fault decision weight coefficient; if both are not within the dynamic determination threshold interval, perform a level determination on the equipment fault and generate a fault level code; Generate a fault trigger value based on the fault decision weight coefficient and the fault decision value, and match it with a preset fault level warning interval to obtain the fault level code; According to the fault level code and the fault duration, the corresponding operation trigger point is activated.

5. The method for generating an interface layout based on an intelligent gateway according to claim 1, characterized in that: The specific steps of determining the interface function weights based on the operation trigger points in combination with a preset weight evaluation model, and constructing the interface layout template in combination with a dynamic layout generation algorithm include: Based on user operation frequency and data update frequency combined with heat maps, the operation interface is divided and high-frequency touch areas are marked; high-frequency touch areas are disassembled according to business processes, key trigger actions are marked, and functional interface areas are divided; According to the business value and information priority, indicators are constructed for the functional interface area to obtain a functional trigger matrix; Responsive rules are set according to the interface size, and the function trigger matrix is ​​iteratively trained in combination with the operation trigger point to obtain a weight evaluation model and output the interface function weight; Allocate all functional areas according to the interface function weights in combination with a dynamic layout generation algorithm to obtain corresponding interface space proportions; According to the interface space proportion and Gestalt principles, all the functional areas are clustered and deployed to obtain an interface layout template.

6. The method for generating an interface layout based on an intelligent gateway according to claim 5, characterized in that: The specific steps of setting responsive rules according to the interface size, iteratively training the function trigger matrix in combination with the operation trigger point to obtain a weight evaluation model and outputting the interface function weight include: Divide the operation interface into preset size groups according to the interface size, define the number of grid columns and component priority; Generate responsive rules based on the number of grid columns and the component priority combined with layout adaptation logic; According to the responsive rules and the operation trigger points, weights are assigned to the function trigger matrix to form a weight vector matrix; According to the preset constraint parameters and the weight vector matrix, the function trigger matrix is ​​iteratively trained several times to obtain a weight evaluation model; Optimizing the initial weight matrix W of the weight evaluation model according to the target loss function L to obtain the optimal weight matrix; Among them, 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 functional module, s j is the device size weight coefficient, M[i,j] is the function trigger matrix element, λ is the regularization coefficient, and Reg is the regularization term; The operation interface is weighted according to the optimal weight matrix combined with user behavior data, and the interface function weight is output.

7. The method for generating an interface layout based on an intelligent gateway according to claim 1, characterized in that: The specific steps of capturing and encoding user operation behaviors, generating interface operation events, and modifying the interface layout template in combination with event processing logic to obtain the optimal interface layout include: Capture and count user operation behaviors on the operation interface to obtain interface operation events; Filter all the interface operation events for invalid operations according to the event monitoring mechanism to obtain valid operation events; Classifying and coding the valid operation events to obtain unique event codes; Sort the functional areas according to the unique event code and event processing logic to generate a function triggering process; The function triggering process is disassembled and associated with the functional modules of the device to obtain a process mapping relationship table; according to the process mapping relationship table and the user's operating habits, the interface layout template is modified to obtain the optimal interface layout.

8. An intelligent gateway, characterized in that: include: The sensor monitoring module is used to display the device deployment environment parameter data collected by sensors in real time; AI recognition module, used to analyze and identify pointer instruments 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 equipment; Status monitoring module, used to monitor the operating status and abnormal alarm information of the smart gateway; Video surveillance module, used to display the station house video stream and remote instrument monitoring screen in real time; Intelligent inspection module, used to automatically perform inspection tasks and provide feedback on equipment operating status; AI analysis module, used to integrate multi-source data to achieve fault prediction and intelligent decision-making analysis; Linkage control module, used to trigger equipment control and emergency response operations according to preset logic; The historical query module is used to store and review historical data and statistical analysis results.

9. An electronic device, characterized in that: include: one or more processors; a memory for storing 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 according to any one of claims 1 to 7.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images