Urban intelligent Internet of Things-based global management method and system

Abstract

The invention relates to the technical field of intelligent city Internet of Things systems, in particular to a global management method and system based on city intelligent Internet of Things, and the method comprises the following steps: collecting equipment operation state data and city business original data; analyzing the historical fault mode and the current operation index to predict the equipment fault risk; adjusting traffic control algorithm parameters and testing effects, and generating a quantitative influence comparison table of parameter changes on traffic efficiency and safety indexes; and performing error diagnosis on the whole process of data acquisition, transmission and analysis, and feeding back a correction instruction to the step of acquiring the operation state data of the equipment and the original data of the urban business. According to the invention, the sensing device and the edge computing node in the urban Internet of Things are connected through the uniform interface, the operation state data of the collected device and the original data of the urban business are fused, and the device fault prediction and the traffic control parameter optimization processing are respectively carried out. And the fusion management and control capability of the urban intelligent Internet of Things system and the accuracy and real-time performance of service decision are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart city Internet of Things systems, and in particular to a global management method and system based on a smart city Internet of Things. BACKGROUND

[0002] In the field of smart city Internet of Things, how to achieve efficient collaborative management of a large number of sensing devices and business systems is a long-term challenge faced by the industry. The current typical technical architecture is mainly divided into two categories: one adopts a centralized data platform to upload the running state data collected by devices such as cameras and radars and city business data such as traffic flow to the cloud for processing; the other relies on edge computing nodes for distributed processing to complete data analysis on the device side.

[0003] In the prior art, the above-mentioned technologies have inherent defects: the time delay and distortion problem of multi-source data fusion decision cannot be effectively solved. The centralized processing architecture causes significant response delay due to remote data transmission and center computing bottleneck. When the device sensor has zero drift or data packet loss, the cloud cannot timely detect the abnormality, resulting in lagging fault prediction and even misjudgment; while the edge computing scheme reduces the delay, but because the device running monitoring and traffic optimization and other business systems independently process data, data islands are formed. More seriously, when the edge node detects device state abnormalities, it can only trigger local alarms but cannot synchronize the data source of the business decision layer, resulting in continuous use of distorted data for traffic parameter adjustment. This fragmented data processing link not only causes passive response in operation and maintenance, such as device downtime followed by maintenance, but also makes business optimization fall into a manual trial and error cycle, engineers need to repeatedly adjust signal light parameters but lack quantitative basis, ultimately leading to lagging and inaccurate city management decisions.

[0004] Therefore, there is an urgent need for a new global management method and system based on a smart city Internet of Things to solve the time delay and distortion problem of multi-source data fusion decision of city Internet of Things. SUMMARY

[0005] In view of at least one of the above technical problems, the present application provides a global management method and system based on a smart city Internet of Things, which adopts a unified interface to fuse device monitoring and business data streams, triggers a closed-loop correction mechanism through double-path analysis and full-process error diagnosis to solve the time delay and distortion problem of multi-source data fusion decision.

[0006] According to a first aspect of the present application, a global management method based on a smart city Internet of Things is provided, comprising the following steps:

[0007] Connecting sensing devices and edge computing nodes in the city Internet of Things through a unified interface to collect device running state data and city business raw data in real time;

[0008] Based on the device operation state data, historical failure modes are analyzed with current operation indexes to predict device failure risk, and when the risk value exceeds a set threshold, an operation and maintenance work order is automatically triggered;

[0009] Meanwhile, based on the city business original data, traffic control algorithm parameters are adjusted and effects are tested to generate a quantitative influence contrast table of parameter changes on traffic efficiency and safety indexes;

[0010] Based on the processing of the device operation state data and the city business original data, error diagnosis is performed on the whole process of data collection, transmission and analysis, when data anomalies are detected, optimization strategies are simulated and verified in combination with the quantitative influence contrast table, and device configuration correction instructions are generated, and the correction instructions are fed back to the steps of collecting device operation state data and city business original data to form a closed loop.

[0011] In some embodiments of the application, the perception device includes at least two of a traffic signal, a camera, and a radar sensor, and the edge computing node performs data cleaning and timestamp alignment preprocessing on the original data.

[0012] In some embodiments of the application, the prediction of device failure risk includes: inputting the historical failure mode into an LSTM model to train a failure predictor, and calculating the deviation degree of the current operation index and the predictor output in real time as the risk value; wherein the operation index includes device temperature, data packet loss rate, and power supply fluctuation amplitude.

[0013] In some embodiments of the application, the set threshold is a dynamic threshold, which is dynamically adjusted according to device type, seasonal factors and historical false alarm rate, and is realized by weighted calculation of device factory reference threshold and past set time fault-free operation probability.

[0014] In some embodiments of the application, the adjustment of the traffic control algorithm parameters includes: iteratively optimizing MPC controller parameters in a digital twin environment based on a reinforcement learning model;

[0015] The quantitative influence contrast table stores the mapping relationship between parameter combinations and traffic efficiency and safety indexes in the form of a matrix.

[0016] In some embodiments of the application, the error diagnosis performed on the whole process of data collection, transmission and analysis includes the following steps:

[0017] When the data of the camera or the radar sensor has a variance that exceeds a set threshold corresponding to the device type within a continuous collection period, data jump diagnosis is triggered;

[0018] When the time delay of the traffic signal state data transmitted to the edge computing node exceeds the set threshold, transmission link diagnosis is triggered;

[0019] The simulation verification strategy is executed based on the quantitative influence control table called according to the diagnosis result.

[0020] In some embodiments of the present application, the generated device configuration correction instruction comprises:

[0021] The camera generates white balance calibration instructions and exposure parameter reset instructions for triggering data jump diagnosis;

[0022] The radar sensor generates transmission power compensation instructions and filter coefficient adjustment instructions for triggering data jump diagnosis;

[0023] The traffic light generates state reset instructions and communication module restart instructions for triggering transmission link diagnosis.

[0024] In some embodiments of the present application, the step of feeding back the correction instruction to the collection device operation state data and urban business original data comprises:

[0025] The device control interface of the edge computing node executes instruction writing: the camera correction instruction is written into its processing unit through a preset video protocol, the radar correction instruction is written into a signal processor through a preset vehicle bus protocol, and the traffic light correction instruction is mapped to an industrial control interface;

[0026] After the instruction is executed, the operation state data of the corresponding device is re-collected.

[0027] In some embodiments of the present application, the edge computing node adopts a multi-thread architecture to execute:

[0028] The data collection thread polls the sensing device interface in real time;

[0029] The model processing thread synchronously executes the operation of predicting the device failure risk and the operation of adjusting the traffic control algorithm parameters;

[0030] The diagnosis monitoring thread monitors data anomalies and transmission delays in real time, and triggers the simulation verification optimization strategy when detecting data anomalies.

[0031] According to the second aspect of the present application, a global management system based on urban intelligent Internet of Things is also provided, comprising:

[0032] The sensing device module comprises at least two of a traffic light, a camera, and a radar sensor, and is used to collect device operation state data and urban business original data;

[0033] The edge computing node module is connected with the sensing device module through a unified interface, and is used to perform data cleaning and time stamp alignment processing on the original data;

[0034] A fault prediction module configured in the edge computing node module is used for inputting historical fault modes into an LSTM model to train a fault predictor, calculating a risk value in real time based on current operation indexes, and automatically triggering an operation and maintenance work order when the risk value exceeds a set threshold dynamically adjusted according to device types, seasonal factors and historical false alarm rates;

[0035] A traffic control optimization module is used for iteratively optimizing MPC controller parameters in a digital twin environment based on urban business raw data through a reinforcement learning model, and generating a mapping relationship between parameter combinations and traffic efficiency and safety indexes in a matrix form as a quantitative influence control table.

[0036] An abnormality diagnosis module is used for triggering data jump diagnosis when the variance of camera or radar data exceeds a device type set threshold in a continuous collection period, or triggering transmission link diagnosis when the transmission delay of a traffic signal lamp exceeds a threshold, and calling the quantitative influence control table to perform strategy simulation verification.

[0037] A device correction instruction module is used for generating device configuration correction instructions according to diagnosis results, including white balance calibration instructions, exposure parameter reset instructions, transmit power compensation instructions, filter coefficient adjustment instructions, state reset instructions and communication module restart instructions.

[0038] An instruction feedback module is used for feeding back correction instructions to sensing devices through a device control interface of the edge computing node, wherein camera correction instructions are written into the processing unit thereof through a preset video protocol, radar correction instructions are written into a signal processor through a preset vehicle bus protocol, and traffic signal lamp correction instructions are mapped to an industrial control interface; the running state data of the corresponding device is re-collected after the execution of the instructions, forming a closed loop.

[0039] A multi-thread scheduling module is used for parallel execution of data collection threads, model processing threads and diagnosis monitoring threads in the edge computing node, so as to realize real-time cooperation of fault risk prediction, traffic parameter optimization and abnormality diagnosis.

[0040] The present application has the following advantages: the present application connects sensing devices and edge computing nodes in the urban Internet of Things through a unified interface, fuses device running state data and urban business raw data, and respectively develops device fault prediction and traffic control parameter optimization processing, thereby constructing a global management mechanism for cooperative operation of devices and business, and realizing active early warning of fault risks, quantitative optimization of traffic strategies, and full-process data abnormality diagnosis and closed-loop correction on the edge side. BRIEF DESCRIPTION OF DRAWINGS

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating the steps of the city-wide smart Internet of Things-based comprehensive management method in an embodiment of the present invention.

[0043] Figure 2 This is a flowchart illustrating the steps involved in error diagnosis throughout the entire data acquisition, transmission, and analysis process in this embodiment of the invention.

[0044] Figure 3 This is a flowchart illustrating the steps for generating device configuration correction instructions in an embodiment of the present invention;

[0045] Figure 4 This is a flowchart illustrating the steps of feeding back correction instructions to the data acquisition device's operating status and the city's original business data in an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0047] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0049] like Figures 1 to 4 The illustrated method for comprehensive management based on the city's smart Internet of Things includes the following steps:

[0050] Through a unified interface, connect the sensing devices and edge computing nodes in the city's Internet of Things, and collect real-time device operation status data and city business raw data;

[0051] Based on the device operation status data, analyze historical failure patterns and current operation indicators to predict device failure risks, and automatically trigger maintenance work orders when the risk value exceeds the set threshold;

[0052] At the same time, based on the city business raw data, adjust the traffic control algorithm parameters and test the effect, and generate a quantitative impact comparison table of parameter changes on traffic efficiency and safety indicators;

[0053] Based on the processing of device operation status data and city business raw data, perform error diagnosis on the whole process of data collection, transmission and analysis, when detecting data anomalies, combine the quantitative impact comparison table to simulate and verify the optimization strategy and generate device configuration correction instructions, and feed back the correction instructions to the steps of collecting device operation status data and city business raw data to form a closed loop.

[0054] For example, Figure 1As shown, in the working process, the urban intelligent Internet of Things platform integrates the sensing devices deployed on the nodes of urban roads through a unified interface, including cameras, radar sensors, traffic signal lights, etc., and establishes a communication connection with the edge computing nodes. The system collects device running state data such as temperature, voltage, current fluctuation, etc., and urban business raw data such as traffic flow, signal control state, etc., in real time, and transmits them to the edge nodes for preprocessing. When the collected device running state data shows that some operation indicators show an abnormal trend, the system analyzes the potential fault risk of the device by comparing the historical operation mode with the current state; when the risk exceeds the preset threshold standard, it automatically triggers the operation and maintenance response process and generates the corresponding processing instructions, which are submitted to the maintenance scheduling system for execution. At the same time, for the collected urban business raw data, the system adjusts the parameters of the traffic control strategy, and evaluates the adjustment results through on-site feedback to form the corresponding relationship data between parameter changes and urban traffic efficiency and safety indicators, which are used to further optimize the traffic control strategy. The system also includes an abnormal monitoring mechanism for the entire data processing process. When conditions such as image signal jump, data variance fluctuation anomaly, or signal light state data transmission delay anomaly are detected, the system starts the error diagnosis process, verifies the optimization strategy in combination with the aforementioned traffic parameter adjustment effect comparison results, and generates device configuration correction instructions based on this. The correction instructions will be fed back to the sensing devices through the edge computing nodes and written into the corresponding device control interface to complete parameter adjustment and configuration update, and then the system automatically triggers the re-collection of the running state data of the device, forming a data closed-loop process from abnormal detection, strategy verification to device repair, realizing the collaborative optimization of devices and business systems. In this embodiment, the present application connects the sensing devices and edge computing nodes in the urban Internet of Things through a unified interface, fuses the collected device running state data and urban business raw data, and respectively develops device fault prediction and traffic control parameter optimization processing, building a global management mechanism for the collaborative operation of devices and business, realizing active early warning of fault risk, quantitative optimization of traffic strategy, and data anomaly diagnosis and closed-loop correction in the edge side. Compared with the prior art, the problems of large data transmission delay and abnormal response lag in centralized processing mode, and the problems of separation of device monitoring and business optimization in edge computing architecture, data distortion and inability to link and correct are avoided, and the fusion management capability and the accuracy and real-time performance of business decision of the urban intelligent Internet of Things system are enhanced.

[0055] In the embodiment of the present application, the perception device includes at least two of traffic lights, cameras, radar sensors, and the edge computing node performs data cleaning and timestamp alignment preprocessing on the raw data. The perception device includes at least two of traffic lights deployed at urban intersections, cameras for image acquisition, and radar sensors for target detection. The above-mentioned perception device is connected with the edge computing node through a unified communication interface, and constitutes part of the urban intelligent Internet of Things perception layer. Specifically, the camera is used to acquire image information such as vehicle flow and pedestrian state; the radar sensor is used to acquire physical state parameters such as the distance and speed of the vehicle; and the traffic light provides its current signal state and switching timing information. The above-mentioned devices can be used in combination according to actual deployment needs, for example, cameras and radars, or cameras and traffic lights, etc. are deployed at the same time, to meet the fusion acquisition needs of multi-source perception information. After receiving the data from the perception device, the edge computing node performs preprocessing operations on the raw data. The preprocessing includes data cleaning of possible null values and noise values in the raw data, and timestamp alignment processing of data collected by different devices. The timestamp alignment operation is used to solve the problem of data asynchronization caused by inconsistent collection periods or clock deviation of different types of devices, to ensure that subsequent analysis and processing are based on a unified time reference, thereby improving the effectiveness and timeliness of data fusion. In this embodiment, standardized and synchronized acquisition and preprocessing of multi-type perception device data are realized, which provides an input basis for subsequent device operation state analysis, traffic control optimization and abnormal diagnosis operations.

[0056] Further, the prediction of the device failure risk comprises: inputting historical failure patterns into an LSTM model to train a failure predictor, and calculating a deviation degree between current operation indexes and the predictor output in real time as the risk value; wherein the operation indexes comprise device temperature, data packet loss rate, and power supply fluctuation amplitude. The system pre-establishes a historical failure pattern dataset containing a large number of known device failure cases, and trains an LSTM model, i.e., a long short-term memory neural network, based on the dataset, thereby constructing a failure predictor. In actual operation, the system inputs current operation indexes from the field sensing device into the trained LSTM failure predictor, and the predictor outputs expected normal operation state parameters of the current device. Subsequently, the system compares the deviation degree between the current actual operation indexes and the prediction output result as the failure risk value of the device. The greater the deviation degree, the greater the difference between the current state of the device and the normal state judged by the historical learning model, and the higher the risk of failure. In this embodiment, the operation indexes at least include three types of parameters, i.e., device temperature, data packet loss rate, and power supply fluctuation amplitude. Among them, the temperature reflects the working load and cooling state of the internal elements of the device, the data packet loss rate reflects the stability of the communication link, and the power supply fluctuation amplitude reflects the quality of the power input of the device. The above indexes can be uploaded in real time by devices such as cameras, radars, and traffic signal lights through their internal sensors, and the edge computing nodes dynamically monitor and analyze them. By introducing the failure prediction mechanism based on the LSTM model, potential abnormal operation trends can be identified in advance before the actual failure of the device, thereby realizing active operation and maintenance management of the city sensing device and improving the overall operation stability and maintenance efficiency of the system.

[0057] Preferably, the threshold value is set as a dynamic threshold value, which is dynamically adjusted according to the device type, seasonal factors and historical false alarm rate, and is realized by weighted calculation of the factory reference threshold value and the past set time fault-free operation probability. Specifically, different types of sensing devices have their own corresponding reference risk threshold values at the factory due to differences in structure, function and operating environment. For example, a camera may be more sensitive to temperature, while a radar is more susceptible to power fluctuations. Therefore, the system first calls the corresponding factory reference threshold value of the device type as the initial reference value. At the same time, the system will combine the seasonal factors of the current operating cycle, such as high temperature in summer and low temperature in winter, which may cause fluctuations in device performance, to dynamically correct the threshold level of temperature-related indicators. In addition, the system continuously records the operating state of the device in the past set period of time, and calculates the probability of fault-free operation in that period of time as a reflection of the stability of the device. Finally, the system generates a dynamic threshold value for risk judgment applicable to the current period by weighted calculation of the factory reference threshold value and the fault-free operation probability in that period. The weighting factor can be adjusted according to experience or set rules to balance the risk between false alarms caused by excessive sensitivity and missed alarms caused by excessive leniency. The use of the above dynamic threshold mechanism makes the fault prediction process more adaptive and stable, effectively reduces the misjudgment caused by environmental disturbance or device characteristic difference, and improves the accuracy and timeliness of operation and maintenance decision.

[0058] In an embodiment of the present application, adjusting the traffic control algorithm parameters includes: iteratively optimizing the MPC controller parameters in the digital twin environment based on the reinforcement learning model; and quantifying the influence table to store the mapping relationship between the parameter combinations and the traffic efficiency and safety indicators in matrix form. For the collected urban business raw data, the system optimizes and adjusts the control parameters in the traffic control algorithm to improve the traffic efficiency and traffic safety indicators. Specifically, the edge computing node or the backend control center deploys a reinforcement learning model for training and optimizing the control strategy in the digital twin environment. The digital twin environment is a simulation platform highly consistent with the actual road traffic environment, which can access data streams from urban perception devices in real time and simulate various vehicles, traffic lights, pedestrians and traffic scenario changes. In this simulation environment, the system uses the MPC (Model Predictive Control) controller as the execution core of the traffic control algorithm, and iteratively optimizes the MPC controller parameters through the reinforcement learning algorithm. In each round of training process, the system uses the traffic efficiency obtained through simulation, such as vehicle traffic speed, waiting time, etc., and the traffic safety indicators, such as conflict rate, emergency braking frequency, etc., as the reward function feedback for evaluating the control effect of different parameter combinations. After multiple iterations, the system converges to one or more sets of relatively optimal parameter combinations. In order to facilitate subsequent quick search and comparison, the system records the mapping relationship between each parameter combination and its corresponding traffic efficiency and safety indicators in the quantitative influence table. The table is stored in matrix form, and each row of the matrix corresponds to a parameter combination, and each column corresponds to a performance indicator, forming a structured data set of parameter optimization and evaluation results. In this embodiment, the dynamic optimization and adaptive adjustment capability of the traffic control algorithm parameters is realized, so that the system can automatically select and match the control strategy according to different intersections, time periods and traffic states, thereby improving the overall operation efficiency and traffic safety level of urban roads.

[0059] In an embodiment of the present application, as shown in Figure 2 Error diagnosis is performed on the whole process of data collection, transmission and analysis, including the following steps:

[0060] When the variance of the data of the camera or radar sensor exceeds the set threshold value corresponding to the device type in the continuous collection period, data jump diagnosis is triggered;

[0061] When the time delay of the traffic signal light state data transmitted to the edge computing node exceeds the set threshold value, transmission link diagnosis is triggered;

[0062] Based on the diagnosis result, the strategy simulation verification is performed based on the quantitative influence table.

[0063] To achieve full-process abnormality identification of the data processing link, the system sets an error diagnosis mechanism in the edge computing node, which is used for real-time monitoring and judgment of the data collection, transmission and analysis process. Specifically, the system continuously monitors the data collected by the camera and radar sensor, and calculates the data variance in a plurality of consecutive collection periods. When the variance of a certain type of device exceeds the set threshold value corresponding to the device type, the system determines that the data has a risk of jumping and triggers data jump diagnosis. The set threshold value can be pre-set according to different types of sensing devices, for example, the image gray scale change amplitude threshold for image devices, or the distance fluctuation threshold for radar devices.

[0064] At the same time, the system also monitors the communication delay of the state data of the traffic signal lamp in the process of transmission to the edge computing node. When the transmission delay exceeds the set threshold value, it means that there is a potential communication abnormality or link blockage, and the system triggers transmission link diagnosis.

[0065] On the basis of completing the above diagnosis, the system further calls the established quantitative influence comparison table according to the diagnosis result to simulate and verify the business strategy that may be affected by the abnormality. In the simulation process, the system compares the traffic efficiency and safety indicators of the parameters under abnormal data conditions with those corresponding to normal parameters, so as to judge whether the control strategy needs to be adjusted or the device correction instruction needs to be generated, to support the subsequent optimization process. In this embodiment, the system can realize dynamic monitoring and classification diagnosis of the data quality of the sensing device and the state of the data link, and provide pre-support for the full-process closed-loop control.

[0066] Specifically, as shown in Figure 3 the generation of device configuration correction instructions includes:

[0067] generating a white balance calibration instruction and an exposure parameter reset instruction for the camera triggering data jump diagnosis;

[0068] generating a transmission power compensation instruction and a filter coefficient adjustment instruction for the radar sensor triggering data jump diagnosis;

[0069] generating a state reset instruction and a communication module restart instruction for the traffic signal lamp triggering transmission link diagnosis.

[0070] When the system identifies a specific type of abnormal event in the error diagnosis process, corresponding device configuration correction instructions are automatically generated for different types of abnormal events of different sensing devices to quickly correct and restore the performance state of the device. Specifically, when the variance of the camera data in the continuous acquisition period exceeds the set threshold and is determined by the system to have a data jump, the system will generate a white balance calibration instruction and an exposure parameter reset instruction. The white balance calibration instruction is used to readjust the color balance in the camera image acquisition process to avoid image color cast caused by changes in ambient light; the exposure parameter reset instruction resets the exposure time or exposure intensity parameter to restore the image brightness to the normal range, improving the stability and usability of the image data. If the radar sensor data is identified as a jump anomaly during diagnosis, the system generates a transmit power compensation instruction and a filter coefficient adjustment instruction. The transmit power compensation instruction is used to modify the energy of the radar transmit signal according to the current working state to ensure that the detection range and sensitivity are within a reasonable range; the filter coefficient adjustment instruction optimizes the parameters of the noise filtering logic in the signal processing module to improve the recognition ability of the effective echo and reduce the probability of misjudgment. For traffic signal light communication abnormalities found in the transmission link diagnosis, if the state data transmission to the edge computing node has a delay exceeding the threshold, the system will issue a state reset instruction and a communication module restart instruction. The state reset instruction is used to reset the current logic state of the signal light to a safe default value to ensure the control stability of the traffic site; the communication module restart instruction restarts the communication unit module of the signal light to attempt to restore its normal data reporting function. In this embodiment, different device types can obtain differentiated and targeted correction instructions in a timely manner after the occurrence of corresponding abnormal events, enhancing the system's automated recovery capability and operational stability and laying the foundation for subsequent control optimization and closed-loop feedback.

[0071] Further, the step of feeding the correction instructions to the acquisition device running state data and the city business raw data includes:

[0072] Write instructions through the device control interface of the edge computing node: camera correction instructions are written to the processing unit through a preset video protocol, radar correction instructions are written to the signal processor through a preset vehicle bus protocol, and traffic signal light correction instructions are mapped to an industrial control interface;

[0073] After the instructions are executed, the running state data of the corresponding device is reacquired.

[0074] As Figure 4As shown, after the system generates the device configuration correction instruction of the camera, radar sensor or traffic signal lamp, the instruction needs to be issued and executed through the edge computing node to realize the parameter configuration adjustment and state recovery of the related device. Specifically, the system completes the write operation of different types of correction instructions through the device control interface of the edge computing node. For the camera, the system encapsulates the generated white balance calibration instruction and exposure parameter reset instruction into a control data packet conforming to its device communication specification, and writes it to the processing unit of the camera through the preset video protocol, so that the internal image processing logic module updates the relevant parameters in time. For the radar sensor, the system converts the generated transmit power compensation instruction and filter coefficient adjustment instruction into a data format compatible with its communication interface, and writes it to the signal processor of the radar through the preset vehicle bus protocol, to ensure that the radar transmits and processes signals according to the updated parameters in the next collection period. For the traffic signal lamp, the system maps the state reset instruction and communication module restart instruction to the standard command format of the industrial control interface according to the mapping rule established in advance, and completes the instruction transmission through the interface to make the state logic and communication module of the signal lamp restart or reset as needed. After the various correction instructions are written through the control interface and successfully executed, the system immediately triggers the operation of reacquiring the running state data of the corresponding device. The reacquired data will overwrite the original abnormal data cache and serve as an important basis for subsequent judgment of whether the correction is completed, ensuring that the correction action is real and effective and realizing the system closed-loop control.

[0075] In the embodiments of the present application, the edge computing node adopts a multi-thread architecture to execute:

[0076] The data acquisition thread polls the sensing device interface in real time;

[0077] The model processing thread synchronously executes the operation of predicting the device failure risk and the operation of adjusting the traffic control algorithm parameters;

[0078] The diagnosis monitoring thread monitors the data anomaly and transmission delay in real time, and triggers the simulation verification optimization strategy when detecting the data anomaly.

[0079] To achieve parallel operation and efficient response of various processing tasks in the urban intelligent Internet of Things system, the edge computing node adopts a multi-thread architecture to perform various key functions. Specifically, the system sets up a data collection thread for interfacing with connected sensing device interfaces, and uses a polling method to obtain real-time device operating state data and urban business raw data, including traffic lights, cameras, and radar sensors. This thread is responsible for maintaining continuous communication with the sensing layer devices to ensure the continuity and real-time nature of data collection. The system also sets up a model processing thread that synchronously executes multiple data analysis-based tasks within the edge node, including: on the one hand, processing device operating state data to predict device failure risks; on the other hand, processing urban business raw data to adjust traffic control algorithm parameters. Both processes are based on the same data input and serve the operation and maintenance warning and traffic optimization goals, respectively, and are completed in parallel in this thread. In addition, the system also sets up a diagnostic monitoring thread that is specifically used to continuously monitor the stability of various types of data and the state of the transmission link. This thread will detect data fluctuations from cameras or radars, as well as the delay of traffic light state data transmission to the edge node in real time. When the detection results indicate that there are data abnormalities, the thread immediately triggers simulation verification optimization strategies to assess whether parameters need to be corrected or device configuration instructions need to be issued. By using the above multi-thread architecture, the system achieves parallelization of data collection, model processing, and abnormality diagnosis at the edge computing node level, improving overall processing efficiency and real-time response capability.

[0080] In an embodiment of the present application, a global management system based on urban intelligent Internet of Things is also provided, comprising:

[0081] A sensing device module, including at least two of traffic lights, cameras, and radar sensors, for collecting device operating state data and urban business raw data;

[0082] An edge computing node module connected to the sensing device module through a unified interface for data cleaning and timestamp alignment of raw data;

[0083] A fault prediction module configured in the edge computing node module for inputting historical fault patterns into an LSTM model to train a fault predictor, calculating a risk value in real time based on current operating indicators, and automatically triggering an operation and maintenance work order when the risk value exceeds a set threshold dynamically adjusted according to device type, seasonal factors, and historical false alarm rate;

[0084] A traffic control optimization module for iteratively optimizing MPC controller parameters in a digital twin environment based on urban business raw data through a reinforcement learning model, and generating a mapping relationship between parameter combinations and traffic efficiency, safety indicators in matrix form as a quantitative impact control table;

[0085] Anomaly diagnosis module, for triggering data jump diagnosis when the variance of camera or radar data exceeds the device type set threshold in the continuous acquisition cycle, or triggering transmission link diagnosis when the traffic signal lamp transmission delay exceeds the threshold, and calling the quantitative impact table to perform strategy simulation verification;

[0086] Device correction instruction module, for generating device configuration correction instructions according to the diagnosis results, including white balance calibration instructions, exposure parameter reset instructions, transmit power compensation instructions, filter coefficient adjustment instructions, state reset instructions and communication module restart instructions;

[0087] Instruction feedback module, for feeding back the correction instructions to the sensing device through the device control interface of the edge computing node, wherein the camera correction instructions are written into the processing unit through the preset video protocol, the radar correction instructions are written into the signal processor through the preset vehicle bus protocol, and the traffic signal lamp correction instructions are mapped to the industrial control interface; the running state data of the corresponding device is reacquired after the execution of the instructions, forming a closed loop;

[0088] Multi-thread scheduling module, for parallel execution of data acquisition threads, model processing threads and diagnosis monitoring threads in the edge computing node, to realize real-time collaboration of fault risk prediction, traffic parameter optimization and anomaly diagnosis.

[0089] Those skilled in the art should understand that the present application is not limited to the above embodiments, and the above embodiments and descriptions in the specification are only to illustrate the principles of the present application, and various changes and improvements can be made without departing from the spirit and scope of the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for comprehensive management based on urban smart IoT, characterized in that, Includes the following steps: Connect sensing devices and edge computing nodes in the city's Internet of Things through a unified interface to collect real-time data on device operating status and raw data of urban business. Based on the equipment operating status data, historical failure modes and current operating indicators are analyzed to predict equipment failure risks. When the risk value exceeds the set threshold, an operation and maintenance work order is automatically triggered. Simultaneously, based on the original data of the city's business, the traffic control algorithm parameters were adjusted and the effects were tested, generating a quantitative impact comparison table on traffic efficiency and safety indicators of parameter changes; Based on the processing of the equipment operation status data and the original urban business data, error diagnosis is performed on the entire process of data collection, transmission and analysis. When data anomalies are detected, the optimization strategy is verified by simulation in conjunction with the quantitative impact comparison table and an equipment configuration correction instruction is generated. The correction instruction is fed back to the steps of collecting equipment operation status data and original urban business data to form a closed loop.

2. The method for comprehensive management based on urban smart IoT according to claim 1, characterized in that, The sensing devices include at least two of traffic lights, cameras, and radar sensors, and the edge computing nodes perform data cleaning and timestamp alignment preprocessing on the raw data.

3. The method for comprehensive management based on urban smart IoT according to claim 2, characterized in that, The method for predicting equipment failure risk includes: inputting the historical failure modes into an LSTM model to train a failure predictor, and calculating in real time the deviation between the current operating indicators and the predictor output as a risk value; wherein the operating indicators include equipment temperature, data packet loss rate, and power supply fluctuation amplitude.

4. The method for comprehensive management based on urban smart IoT according to claim 3, characterized in that, The set threshold is a dynamic threshold, which is dynamically adjusted according to the equipment type, seasonal factors and historical false alarm rate. It is achieved by weighting the equipment's factory benchmark threshold with the probability of fault-free operation over a set period in the past.

5. The method for comprehensive management based on urban smart IoT according to claim 1, characterized in that, The adjustment of traffic control algorithm parameters includes: iteratively optimizing MPC controller parameters in a digital twin environment based on a reinforcement learning model; The quantitative impact comparison table stores the mapping relationship between parameter combinations and traffic efficiency and safety indicators in matrix form.

6. The method for comprehensive management based on urban smart IoT according to claim 2, characterized in that, The error diagnosis process for the entire data acquisition, transmission, and analysis workflow includes the following steps: When the variance of the data from the camera or the radar sensor exceeds the set threshold corresponding to the device type within a continuous acquisition period, a data jump diagnosis is triggered. When the delay in transmitting the traffic light status data to the edge computing node exceeds the set threshold, a transmission link diagnosis is triggered. Based on the diagnostic results, the quantitative impact comparison table is invoked to perform strategy simulation verification.

7. The method for comprehensive management based on urban smart IoT according to claim 6, characterized in that, The generated device configuration correction instruction includes: Generate a white balance calibration command and an exposure parameter reset command for the camera that triggers the data jump diagnosis; The radar sensor that triggers the data jump diagnosis generates a transmit power compensation command and a filter coefficient adjustment command. The traffic light that triggers the transmission link diagnosis generates a status reset command and a communication module restart command.

8. The method for comprehensive management based on urban smart IoT according to claim 7, characterized in that, The step of feeding back the correction command to the data acquisition equipment's operating status data and the city's original business data includes: The device control interface of the edge computing node executes the following instructions: camera correction instructions are written to its processing unit through a preset video protocol, radar correction instructions are written to the signal processor through a preset vehicle bus protocol, and traffic light correction instructions are mapped to the industrial control interface. After the command is executed, the corresponding device's operating status data is re-collected.

9. The method for comprehensive management based on urban smart IoT according to claim 1, characterized in that, The edge computing nodes employ a multi-threaded architecture: The data acquisition thread polls the sensing device interface in real time. The model processing thread synchronously executes the operation of predicting the risk of equipment failure and the operation of adjusting the parameters of the traffic control algorithm; The diagnostic monitoring thread monitors data anomalies and transmission latency in real time, and triggers the simulation verification optimization strategy when a data anomaly is detected.

10. A city-wide smart Internet of Things-based comprehensive management system, characterized in that, include: The sensing device module includes at least two of traffic lights, cameras, and radar sensors, used to collect device operating status data and raw urban business data. The edge computing node module is connected to the sensing device module through a unified interface and is used to perform data cleaning and timestamp alignment on the raw data. The fault prediction module, configured in the edge computing node module, is used to input historical fault patterns into the LSTM model to train the fault predictor, calculate the risk value in real time based on the current operating indicators, and automatically trigger an operation and maintenance work order when the risk value exceeds the set threshold dynamically adjusted according to the equipment type, seasonal factors and historical false alarm rate. The traffic control optimization module is used to iteratively optimize the MPC controller parameters in a digital twin environment based on the original urban business data and through a reinforcement learning model. It also generates a matrix-based mapping relationship between parameter combinations and traffic efficiency and safety indicators as a quantitative impact comparison table. The anomaly diagnosis module is used to trigger data jump diagnosis when the variance of camera or radar data exceeds the threshold set by the device type within a continuous acquisition period, or to trigger transmission link diagnosis when the transmission delay of traffic lights exceeds the threshold, and to call the quantification impact checklist to perform strategy simulation verification. The equipment correction instruction module is used to generate equipment configuration correction instructions based on diagnostic results, including white balance calibration instructions, exposure parameter reset instructions, transmit power compensation instructions, filter coefficient adjustment instructions, status reset instructions, and communication module restart instructions; The instruction feedback module is used to feed back correction instructions to the sensing device through the device control interface of the edge computing node. The camera correction instructions are written to its processing unit through a preset video protocol, the radar correction instructions are written to the signal processor through a preset vehicle bus protocol, and the traffic light correction instructions are mapped to the industrial control interface. After the command is executed, the corresponding device's operating status data is re-collected, forming a closed loop. The multi-threaded scheduling module is used to execute data acquisition threads, model processing threads, and diagnostic monitoring threads in parallel on the edge computing node to achieve real-time coordination of fault risk prediction, traffic parameter optimization, and anomaly diagnosis.

Images