A unified access management system for traffic Internet of Things sensing devices

Abstract

The invention relates to the technical field of Internet of Things, and in particular to a unified access management system for traffic Internet of Things sensing devices, comprising: an access identification module, used for identifying currently accessed Internet of Things devices, and determining the device type and the number of accessed devices corresponding to the Internet of Things devices; a data transmission module, used for determining the stability of data transmission according to the accessed Internet of Things devices, and obtaining the transmission stability coefficient corresponding to the Internet of Things devices; a device alarm module, used for determining multiple measurement points of the Internet of Things devices, and monitoring whether there are alarms at these measurement points; recording the number and distribution of alarms in each cycle, and calculating the alarm distribution coefficient; adjusting the alarm distribution coefficient according to the calculated value of the alarm distribution coefficient, and using the adjusted alarm distribution coefficient as the output result of the device alarm module; achieving the stability and reliability of the access management system, and improving the efficiency and security of the management of Internet of Things devices.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a unified access management system for traffic Internet of Things sensing devices. Background Art

[0002] Existing technology: The current IoT device access management system has a single access method and a single type of device that can be accessed. The generation of alarms is not flexible enough and has certain limitations. Some IoT management systems can only access a certain type of device or can only connect to a certain operator; some IoT management systems have alarm conditions built into the device and cannot be flexibly modified after production.

[0003] For example, Chinese patent publication number CN110675008A discloses a data quality verification method, verification system and storage medium for an IoT device, which belongs to the field of blockchain technology. The data quality verification method in the present invention is: counting the number of IoT devices connected to the blockchain in a preset area, and obtaining device information of the connected IoT devices; comparing the access number with a preset device access number threshold: if the access number is less than or equal to the access threshold, then comparing the uploaded data with the uploaded data of the preset target IoT device to determine the data quality; if the access number is greater than the access threshold, then randomly selecting multiple IoT devices connected to the blockchain in the preset area, comparing the uploaded data with the uploaded data of the randomly selected multiple IoT devices connected to the blockchain, and determining the data quality.

[0004] However, the existing technology only verifies the data quality after the data is uploaded, and cannot adjust the device according to the working status of the device and the corresponding workload. At the same time, it is also necessary to verify the stability of the device itself to determine how to control and manage the connected device. Summary of the invention

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a unified access management system for traffic Internet of Things sensing devices, including: an access identification module, used to identify the currently connected Internet of Things device, and determine the device type and number of access devices corresponding to the Internet of Things device.

[0006] The data transmission module is used to determine the stability of data transmission based on the connected IoT devices and obtain the transmission stability coefficient corresponding to the IoT devices.

[0007] The device alarm module is used to determine multiple measurement points of the IoT device and monitor whether there are alarms at these measurement points; record the number and distribution of alarms in each cycle and calculate the alarm distribution coefficient; adjust the alarm distribution coefficient according to the calculated value of the alarm distribution coefficient, and use the adjusted alarm distribution coefficient as the output result of the device alarm module.

[0008] The state detection module is used to identify the working state of the IoT device when it is connected, identify the workload corresponding to the current working state of the IoT device, analyze the workload of the IoT device and the number of connected devices in each cycle, and obtain the state recognition coefficient corresponding to the working state.

[0009] The quality verification module is used to determine the working model of the currently connected IoT device based on the output results of the data transmission module, the device alarm module and the status detection module, and manage the connected IoT device through the obtained working model.

[0010] The beneficial effects of the present invention are: 1. The present invention adopts a data transmission module to obtain the packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability, and obtains the transmission stability coefficient by weighted summation after standardized processing, comprehensively evaluates the stability of data transmission, and timely discovers and solves the technical effect of transmission problems, thereby improving the reliability and stability of data transmission.

[0011] 2. The present invention uses an equipment alarm module to monitor whether there are alarms at multiple measurement points, record the number and distribution of alarms in each cycle, calculate the alarm distribution coefficient, and dynamically adjust the alarm distribution coefficient based on historical data; monitor and process the alarm conditions of multiple measurement points in real time to improve the sensitivity and accuracy of the alarm system; and effectively manage and respond to alarm events. The technical effect reduces false alarms and missed alarms and improves the safety and reliability of the system.

[0012] 3. The present invention obtains CPU utilization, memory utilization and network bandwidth utilization through a status detection module, obtains the workload by weighted summation, and establishes a relationship model between workload and the number of access devices through historical data analysis, and calculates a status recognition coefficient; accurately identifies the working status of the device, and calculates a status recognition coefficient that reflects the overall performance of the device; optimizes the technical effect of the operating status of the device, and improves the operating efficiency and stability of the device.

[0013] 4. The present invention uses a quality verification module to extract the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment to form a working model; determines the boundary value of the working model, traverses the boundary conditions corresponding to the boundary value, and selects the boundary condition with the greatest similarity as the output working model; comprehensively evaluates the output results of each module, constructs a comprehensive working model, and improves the overall stability and reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0015] Figure 1It is a system framework diagram of a unified access management system for transportation Internet of Things sensing devices.

[0016] Figure 2 It is a system flow chart of a unified access management system for transportation Internet of Things sensing devices.

[0017] Figure 3 It is a logical schematic diagram of a unified access management system for traffic Internet of Things sensing devices. DETAILED DESCRIPTION

[0018] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. If no specific techniques or conditions are specified in the embodiments, the techniques or conditions described in the literature in the art or the product specifications are used.

[0019] See also Figure 1 , Figure 2 , Figure 3 , a unified access management system for traffic Internet of Things sensing devices, including: access identification module, data transmission module, equipment alarm module, status detection module, and quality verification module.

[0020] The access identification module is used to obtain the current number of access devices and the corresponding device types, and to start the identification of the devices. The data transmission module is used to receive instructions from the access identification module and verify the stability of the output transmission. The device alarm module is connected to the data transmission module, and measurement points are set for inspection during data transmission. The status detection module receives the output of the device alarm module and performs device status detection. The quality verification module completes the management of IoT devices by receiving data from the status detection module and integrating the output results of the data transmission module, the device alarm module and the status detection module.

[0021] The access identification module is used to identify the currently connected IoT device and determine the device type and number of connected devices corresponding to the IoT device.

[0022] The data transmission module is used to determine the stability of data transmission based on the connected IoT devices and obtain the transmission stability coefficient corresponding to the IoT devices.

[0023] The device alarm module is used to determine multiple measurement points of the IoT device and monitor whether there are alarms at these measurement points; record the number and distribution of alarms in each cycle and calculate the alarm distribution coefficient; adjust the alarm distribution coefficient according to the calculated value of the alarm distribution coefficient, and use the adjusted alarm distribution coefficient as the output result of the device alarm module.

[0024] The state detection module is used to identify the working state of the IoT device when it is connected, identify the workload corresponding to the current working state of the IoT device, analyze the workload of the IoT device and the number of connected devices in each cycle, and obtain the state recognition coefficient corresponding to the working state.

[0025] The quality verification module is used to determine the working model of the currently connected IoT device based on the output results of the data transmission module, the device alarm module and the status detection module, and manage the connected IoT device through the obtained working model.

[0026] At this time, when unified access management is performed for IoT devices, the stability and efficiency of transmission are mainly identified by judging the interface of the IoT device when it is connected to the current system, whether the device can perform stable data transmission within the current IoT, and whether there are corresponding alarms for data transmission.

[0027] After completing the verification of data transmission, further judgment is made based on the working status of the device to determine the impact of the working status and the number of connected devices on the overall data transmission. After quantifying this part of the impact, the corresponding quality of the overall IoT device access can be obtained, thereby selecting the number of devices to be connected and whether to access the current device.

[0028] In the access identification module, the type of device accessing the Internet of Things is first determined, and a comparison is made to see whether the current device type can access the current network. The judgment method is to determine whether the device type meets the specifications set by the current system to determine whether it can access.

[0029] Then determine the number of connected devices at this time, and determine whether the number of connected devices at this time exceeds the device quantity threshold. If it does not exceed the device quantity threshold, compare the data transmitted by the device to determine the data quality, and select the device with qualified data quality for connection based on the data quality verified later; verify the data quality by performing a cyclic inspection through the device alarm module, status detection module and quality verification module. At this time, it will be checked whether the connected devices meet the requirements, and the devices that meet the requirements will be connected; if it exceeds the device quantity threshold, a part of the devices will be selected each time for data quality verification until all devices are verified.

[0030] At this time, the device type is classified into cameras, radars, geomagnetic sensors, etc. according to the information reported by the device; the number of connected devices is counted by the number of devices of each type to generate a device list.

[0031] Since this is a device used for the transportation Internet of Things, the focus at this time will be on the data upload interval and the upload speed to determine the stability of the device transmission.

[0032] In one embodiment of the present invention, the method for obtaining the transmission stability coefficient is to obtain the data packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability of the Internet of Things device during data transmission, and standardize the data packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability to obtain the transmission stability coefficient.

[0033] ;in, Expressed as the transmission stability coefficient, represents the packet loss rate, represents the average delay, Indicates the jitter duration. Represents the retransmission rate, Indicates the number of connection interruptions. Indicates availability; represents the weight of the packet loss rate, represents the weight of the average delay, Indicates the weight of the jitter duration, represents the weight of the retransmission rate, Indicates the weight of the number of connection interruptions, Indicates the weight of availability; the weight selected at this time is set in the order of packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability, which are 0.3, 0.1, 0.15, 0.05, 0.2, and 0.2, respectively.

[0034] The above-mentioned data packet loss rate is obtained by recording the number of lost data packets of the connected IoT devices through a Ping test.

[0035] The above average delay refers to the average time required for a data packet to travel from the sender to the receiver, that is, the average time it takes for the current system and the connected IoT devices to send and receive data packets. The round-trip time for each data packet is recorded and then the average value is taken.

[0036] The jitter duration mentioned above refers to the delay difference between consecutive data packets. The delay difference between adjacent data packets, that is, the time difference between the round trip of the data packets, is calculated, and the average value of the time difference is taken as the jitter value at this time.

[0037] The above retransmission rate refers to the proportion of data packets that need to be resent due to transmission errors.

[0038] The above-mentioned number of connection interruptions refers to the number of times the connection between the IoT device and the server is interrupted within a certain period of time.

[0039] The above availability refers to the proportion of time the device is online, that is, the proportion of the online time after the IoT device is connected to the system to the total connection time.

[0040] When verifying the transmission stability coefficient in the data transmission module, it is necessary to divide each connected IoT device into different groups according to the time interval for uploading data and the device type, and then calculate the transmission stability coefficient. If there is no corresponding alarm on the device at this time, the device without alarm will be directly connected, otherwise the alarm of the device will be verified and the corresponding data will be transmitted to the device alarm module.

[0041] In one embodiment of the invention, when setting the measurement point in the device alarm module, first determine the number of devices connected at this time, set the measurement point in proportion to the number of connected devices, the measurement point records data in a period of 900 seconds, records the change in the number of connected devices over time in each period and the corresponding alarm point when the number of connected devices changes, and each measurement point selected at this time represents a device; the alarm point indicates that the current system has found a value that exceeds the safety range at the corresponding time, and the value of the alarm point at this time can be the number of connected devices, data packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions, data transmission time, Data transmission time interval. At this time, the values ​​of these data that exceed the safety range are regarded as alarm points, and the distribution and frequency of alarms are verified based on the alarm points identified in this part. At this time, the number and distribution of corresponding alarms tend to verify the joint probability of the existence of alarm points. The alarm distribution coefficient is obtained based on the joint probability of the alarm points. At this time, the total probability of the alarm points appearing under different conditions can be identified through the joint probability. This probability can reflect the impact of the number of connected devices and the quality of data transmission on the alarm points, so as to determine under what conditions the alarm is likely to occur, so as to facilitate the access control of IoT devices and intervene in the automatic access processing of the system in advance.

[0042] At this time, when verifying that an alarm occurs, a safety range is set for each indicator. Exceeding these thresholds is considered an alarm point, and the upper limit of these safety ranges is considered an alarm condition. For example: number of access devices: [0,100]; packet loss rate: [0,0.05]; average delay: [0,100ms]; jitter duration: [0,10ms]; retransmission rate: [0,0.05]; number of connection interruptions: [0,1 times / cycle]; data transmission time: [0,10s]; data transmission time interval: [0,5s]; at this time, if a value exceeding these data is identified at the measurement point, the corresponding measurement point will be identified as an alarm point, and the individual probability of each alarm point will be recorded, that is, the probability of an alarm occurring in all records.

[0043] The method for obtaining the alarm distribution coefficient is as follows: selecting the measurement points whose values ​​exceed the safety range, and treating the measurement points that exceed the safety range as alarm points; obtaining the alarm point set S, where there are n alarm points in the alarm point set, calculating the alarm probability of each alarm point in the alarm point set and the joint probability of every two alarm points, and obtaining the alarm distribution coefficient based on the calculated alarm probability and joint probability; the alarm probability is expressed as the ratio of the number of alarms at the current alarm point to the total number of alarms, and the joint probability represents the ratio of the number of alarms at any two points in the alarm point set to the total number of alarms.

[0044] ;in, Represents the alarm distribution coefficient of the alarm point set, Indicates the number of alarm points in the alarm point set. represents the weight of the i-th alarm point, represents the alarm probability of the ith alarm point, Represents a tuning parameter used to balance the impact of the probability of a single alarm point and the joint probability. It is usually a positive number less than 1, such as 0.1 or 0.2. Indicates the alarm point set. Indicates the point pair of alarm points. In this case, when there are n alarm points, two alarm points are selected each time to form a point pair, and the point pairs coexist indivual; represents the joint weight of the i-th alarm point and the j-th alarm point, It represents the joint probability of the i-th alarm point and the j-th alarm point, and the value range of i and j is 1 to n.

[0045] In the present invention, after completing the calculation of the alarm distribution coefficient, it is necessary to determine the value of the alarm distribution coefficient to prevent the alarm distribution coefficient of the alarm point from being calculated too large; because at this time, the devices connected to the Internet of Things are of the same model, and the safety range of their corresponding devices may have different range values. At this time, it is necessary to adjust the weights during calculation according to the obtained alarm distribution coefficient to adapt to the different safety ranges of different devices.

[0046] The method for adjusting the alarm distribution coefficient is as follows.

[0047] Verify whether the alarm distribution coefficient calculated at this time is greater than the average value of the alarm distribution coefficient in the historical data. When it is greater than the average value of the alarm distribution coefficient in the historical data, sort the alarm points according to the values ​​of the alarm probability and the joint probability from large to small, and select the alarm points whose alarm probability and joint probability are both in the top 10% and the bottom 10% to form an alarm point adjustment set; calculate the alarm distribution coefficient of the alarm point adjustment set, adjust the weight and joint weight of the alarm point according to the alarm distribution coefficient of the alarm point adjustment set and the alarm distribution coefficient of the alarm point set, adjust the alarm distribution coefficient according to the adjusted alarm point weight and joint weight, and use the adjusted alarm distribution coefficient as the alarm distribution coefficient output by the equipment alarm module.

[0048] If it is less than or equal to the average value of the alarm distribution coefficients in the historical data, the corresponding alarm distribution coefficient is used as the alarm distribution coefficient output by the device alarm module.

[0049] At this time, the data of the alarm points whose alarm probability and joint probability are in the top 10% and bottom 10% after sorting are selected as the adjusted alarm points, and the alarm distribution coefficients are calculated for these alarm points. Adjustments are made based on the changing value of the alarm distribution coefficient to adjust the weights of the alarm points calculated at this time, so as to obtain an alarm distribution coefficient that better reflects the current operation status of the access device.

[0050] At this time, the method of selecting the alarm point adjustment set is, for example, if there is an alarm point in the alarm point set whose alarm probability is in the top 10%, and this alarm point is also in the top 10% when calculating the joint probability with other alarm points, then this alarm point will be selected as an element of the alarm point adjustment set. The purpose of the alarm point adjustment set is to determine the points where obvious changes occur and use these points as the basis for adjustment at this time.

[0051] The implementation method of adjusting the weight of the alarm point is as follows.

[0052] ;in, represents the weight of the i-th alarm point after adjustment, represents the weight of the i-th alarm point, Represents the alarm distribution coefficient of the alarm point set, Indicates the alarm distribution coefficient of the alarm point adjustment set, Represents the average value of the alarm distribution coefficient in historical data, Indicates the number of alarm points in the alarm point set. Indicates the number of alarm points in the alarm point adjustment set. Represents an exponential constant.

[0053] The implementation of adjusting the joint weight is as follows.

[0054] ;in, represents the joint weight of the adjusted i-th alarm point and the j-th alarm point, represents the joint weight of the i-th alarm point and the j-th alarm point, represents the exponential constant, Represents the alarm distribution coefficient of the alarm point set, Indicates the alarm distribution coefficient of the alarm point adjustment set, Indicates the number of alarm points in the alarm point set. Indicates the number of alarm points in the alarm point adjustment set.

[0055] The alarm distribution coefficient at this time is recalculated according to the adjusted joint weight and the weight of the alarm point, and the recalculated alarm distribution coefficient is output to obtain the final alarm distribution coefficient. The output value at this time can be continuously adjusted according to the corresponding value range to enhance the prediction ability of low-probability events existing in the alarm point.

[0056] The working status at this time is mainly to identify the working status during access, such as determining the relationship between the workload of the device and the number of access devices, and according to the relationship between the number of access devices, determining the corresponding state recognition coefficient of the IoT device when it is working.

[0057] For example, when a device fails, whether the device load increases or decreases, or whether the default transmission and load of the device and related access conditions are affected when the device is just started, configured, or offline.

[0058] First, define different working status categories, such as: normal operation: the device operates as expected; standby / sleep: the device temporarily stops data collection; maintenance mode: the device is performing software updates or hardware checks; fault / abnormality: the device detects an internal error or external environment changes that cause it to fail to work properly; initialization / startup: the device is performing self-test and initialization after just powering on or restarting; configuring: the device is receiving new configuration instructions; offline / disconnected: the device loses connection with the central server; low power: for battery-powered devices, the device enters a low power state when the power is below a certain threshold; test mode: the device is in a state of debugging or verifying specific functions; safe mode: the safe mode to which the device switches when encountering serious problems.

[0059] The implementation method of calculating the workload corresponding to the current working state of the IoT device is to obtain the CPU usage, memory usage and network bandwidth usage of the IoT device, and use the weighted sum of the CPU usage, memory usage and network bandwidth usage as the workload corresponding to the current working state of the IoT device.

[0060] The workload is represented as shown below.

[0061] ;in, Expressed as workload, Expressed as CPU usage, Expressed as memory usage, Expressed as network bandwidth usage, Expressed as the weight of CPU usage, Expressed as the weight of memory usage, Indicates the weight of network bandwidth usage. The weights of CPU usage, memory usage, and network bandwidth usage are set to 0.5, 0.2, and 0.2, respectively.

[0062] The above CPU usage is expressed as: Get the current CPU usage of the device. Since most IoT devices are devices such as cameras and sensors, there are a certain number of devices that cannot directly obtain the CPU usage. At this time, the CPU usage tends to be the proportion of the time the IoT device executes work tasks to the total time. For example, the executed work tasks are shown below.

[0063] Video acquisition: Read high-definition video data at 30 frames per second from the image sensor.

[0064] Image processing: Noise reduction and color correction are performed on each frame.

[0065] Coding compression: Use the H.264 encoder to compress the processed video data.

[0066] Network transmission: The compressed video stream is sent to the monitoring center via the RTSP protocol.

[0067] Motion detection: Analyze video streams in real time to detect whether there are moving objects.

[0068] Logging: Record system operation status and event logs.

[0069] User interface: Provides a web interface for users to configure camera parameters.

[0070] At this time, if the camera is set to high resolution (such as 4K) and high frame rate (such as 60fps), the CPU usage may be very high, possibly reaching 80% or even higher; enabling multiple intelligent analysis functions (such as motion detection, face recognition, etc.) will also significantly increase the CPU load; if the network bandwidth is limited or the network is unstable, the CPU usage may also be affected. At this time, the IoT device will generate different CPU usage according to the set work tasks.

[0071] The above memory usage is expressed as: Get the current memory usage of the device. In this case, the memory usage is expressed as the ratio of the currently used memory to the total available memory.

[0072] The above network bandwidth utilization rate is expressed as: obtaining the current network bandwidth utilization rate of the device, and the network bandwidth utilization rate is the ratio of the bandwidth currently used by the IoT device to the overall available bandwidth.

[0073] The implementation method for analyzing the workload and the number of access devices of IoT devices in each cycle is to establish a relationship model between the workload and the number of access devices through statistical analysis of historical data; the relationship model set at this time will determine the relationship between the workload and the number of access devices by means of time series and regression analysis, and complete the processing of the working status in combination with the transmission stability coefficient at this time to obtain the state recognition coefficient.

[0074] At this time, the processing method for the relationship model is expressed as follows: divide the workload into multiple time points according to the set period, normalize the workload and the number of access devices, calculate the covariance of the workload and the number of access devices at the corresponding time points, the standard deviation of the number of access devices, and the average value of the workload and the number of access devices, and obtain the relationship model between the workload and the number of access devices.

[0075] The relationship model between workload and the number of connected devices can be expressed as shown below.

[0076] ;in, Expressed as workload, Indicates the number of connected devices. represents the slope of the relationship model, Represents the intercept of the relationship model.

[0077] The slope of the relationship model at this time It can be expressed as shown below.

[0078] ;in, represents the covariance of workload and number of connected devices, Indicates the standard deviation of the number of connected devices.

[0079] The covariance between workload and number of connected devices can be expressed as .

[0080] ;in, represents the workload at the tth time point, represents the average value of the workload, Indicates the number of time points, the value of t ranges from 1 to m; represents the number of devices connected at the tth time point, Indicates the average number of connected devices.

[0081] For the intercept of the relationship model at this time It can be expressed as shown below.

[0082] ;in, represents the average value of the workload, Indicates the average number of connected devices.

[0083] At this time, the acquired relationship model is verified to determine the determination coefficient of the relationship model. When the determination coefficient is greater than the standard determination value, the state recognition coefficient is obtained according to the currently calculated workload and the transmission stability coefficient.

[0084] ;in, represents the coefficient of determination, Represents the workload at the tth time point output by the relational model, represents the workload at the tth time point, represents the average value of the workload, Indicates the number of time points, and the value of t ranges from 1 to m.

[0085] When the determination coefficient is 1, it means that the predicted workload is consistent with the actual value. At this time, the selected determination coefficient needs to satisfy a value greater than 0.95 to ensure that the predicted workload is almost consistent with the actual workload performance.

[0086] The state identification coefficient is expressed as follows.

[0087] ;in, represents the state recognition coefficient, represents the weight of the workload, Expressed as the transmission stability coefficient, represents the workload; the weight of the workload selected at this time is 0.8.

[0088] The state recognition coefficient calculated at this time is the state recognition coefficient selected when the relationship model between the workload and the number of access devices satisfies the value of the determination coefficient, indicating the corresponding state recognition coefficient when a specific workload value occurs.

[0089] The cycle of the above workload setting is consistent with the cycle of the measurement point setting. Both obtain data within 900s and process this data. At this time, the workload verification is more inclined to verify the number of devices connected at the time point corresponding to the workload, as well as the total load of the connected devices, to verify the analysis between the workload and the number of connected devices; and the above measurement points count the values ​​of each measurement point in each cycle, focusing on the average value of each device in the set cycle, or the number of measurement points with problems.

[0090] The goal of combining workload and transmission stability coefficient is to verify whether the current IoT devices can maintain normal workload and stable data transmission when the number of connected devices changes, so as to verify whether the working status at this time can meet the system requirements.

[0091] In one embodiment of the present invention, the working model represents the working model at this time composed of the corresponding values ​​of the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the device extracted from the output results. At this time, the corresponding data of the currently accessed device can be described, which is convenient for timely control and verification of the overall status of the Internet of Things device, and timely adjustment according to the situation of the Internet of Things device.

[0092] Then the implementation method of the working model at this time is as follows.

[0093] According to the output results of the data transmission module, interface alarm module and status detection module, the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment are extracted; the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment are combined into a working model, and the boundary value of the working model is determined. The boundary value at this time is expressed as the numerical value corresponding to the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment. The boundary value is to unify these values ​​and retrieve the boundary conditions required for the working model based on these values. The boundary conditions at this time are expressed as device type and specifications, working status, resource usage, data transmission characteristics, alarm conditions, configuration parameters, and log records. Through these boundary conditions, the control method and specific implementation process of each different IoT device during actual access can be determined.

[0094] The boundary value of the working model is used as the retrieval condition, the boundary conditions corresponding to the boundary value are traversed, and the boundary condition with the greatest similarity among the boundary conditions is used as the output working model.

[0095] At this time, the similarity of the boundary conditions is calculated by calculating the cosine similarity between the values ​​corresponding to the working status, state recognition coefficient, alarm distribution coefficient, and transmission stability coefficient of the current device and the average values ​​corresponding to the working status, state recognition coefficient, alarm distribution coefficient, and transmission stability coefficient of the device in all historical data. The calculation formula used is the calculation formula of the Pearson correlation coefficient; the input of the Pearson correlation coefficient calculation formula is two numbers. At this time, the working status, state recognition coefficient, alarm distribution coefficient, and transmission stability coefficient will be normalized and converted into values ​​of the same dimension, and grouped in pairs into six groups; the six groups are the Pearson correlation coefficient of the working status and the state recognition coefficient, the working status, the state recognition coefficient, the alarm distribution coefficient, and the transmission stability coefficient. The Pearson correlation coefficient between the working state and the alarm distribution coefficient, the Pearson correlation coefficient between the working state and the transmission stability coefficient, the Pearson correlation coefficient between the state recognition coefficient and the alarm distribution coefficient, the Pearson correlation coefficient between the state recognition coefficient and the transmission stability coefficient, and the Pearson correlation coefficient between the alarm distribution coefficient and the transmission stability coefficient; the Pearson correlation coefficients between these four parameters are calculated in turn, and the weighted average of the Pearson correlation coefficients calculated for these six groups is taken, and the weighted average is compared with the preset threshold after weighted average. The preset threshold is 0.8, and the boundary condition that is greater than the preset threshold and has the largest value after calculation at this time is used as the output data. This boundary condition is the currently required working model.

[0096] For example, there is a camera at this time, and its working model can be expressed as shown below.

[0097] 1. Device type and specifications: Processor: ARM Cortex-A7, 1.2GHz; Memory: 512MB RAM; Storage: built-in 32GB SD card; Functions: video acquisition, motion detection, face recognition; Communication protocols: HTTP, RTSP, ONVIF, MQTT.

[0098] 2. Working status: Normal operation: collect video and transmit it in real time, and perform intelligent analysis; Standby / sleep: pause video acquisition, but still maintain network connection; Maintenance mode: download and install firmware updates; Fault / abnormal: hardware fault or network interruption is detected; Initialization / startup: self-test and initialization after power-on; Configuring: receive and apply new configuration parameters; Offline / disconnected: lost connection with the monitoring center; Low battery: battery power is less than 20%; Test mode: perform functional testing and debugging; Safe mode: safe mode entered after a system crash.

[0099] 3. Resource usage: CPU usage: usually between 30%-60%, up to 80% under high load; memory usage: usually between 40%-70%; network bandwidth usage: usually between 500kbps-2Mbps, depending on the video resolution and frame rate.

[0100] 4. Data transmission characteristics: Packet loss rate: usually less than 0.1%; Average delay: usually less than 200ms; Jitter duration: usually less than 50ms; Retransmission rate: usually less than 0.05%; Number of connection interruptions: no more than 1 time per hour.

[0101] 5. Alarm conditions: CPU usage exceeds 80%: triggers a serious alarm; memory usage exceeds 90%: triggers a serious alarm; network interruption exceeds 1 minute: triggers an emergency alarm; packet loss rate exceeds 1%: triggers a warning alarm.

[0102] 6. Configuration parameters: Video resolution: 1080p; Frame rate: 30fps; Encoding format: H.264; Intelligent analysis function: Enable motion detection and face recognition; Network settings: Static IP address 192.168.1.100, subnet mask 255.255.255.0, gateway 192.168.1.1; Authentication information: User name admin, password password.

[0103] 7. Log records: Event log: records events such as device startup, shutdown, configuration changes, etc.; Operation log: records user login, configuration modification and other operations; Performance log: records performance indicators such as CPU usage, memory usage, network bandwidth usage, etc. every 5 minutes.

[0104] At this point, the system can collect device information regularly; trigger corresponding alarm notifications and take corresponding measures based on preset alarm conditions; analyze device resource usage, identify bottlenecks and adjust configuration parameters, such as reducing video resolution or frame rate to reduce CPU and network load.

[0105] When a device fails, quickly locate the problem through log records and status information; take appropriate recovery measures, such as restarting the device, updating firmware, etc.; remotely configure device parameters through API or management interface; regularly back up configuration files to ensure rapid recovery in case of device failure; regularly update firmware and fix known vulnerabilities; monitor network traffic to prevent unauthorized access and malicious attacks.

[0106] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention and they are still covered by the protection scope of the present invention.

Claims

1. A unified access management system for traffic Internet of Things sensing devices, characterized in that: include: The access identification module is used to identify the currently connected IoT device and determine the device type and number of connected devices corresponding to the IoT device; The data transmission module is used to determine the stability of data transmission according to the connected IoT devices and obtain the transmission stability coefficient corresponding to the IoT devices; The device alarm module is used to determine multiple measurement points of the IoT device and monitor whether there are alarms at these measurement points; record the number and distribution of alarms in each cycle and calculate the alarm distribution coefficient; adjust the alarm distribution coefficient according to the calculated value of the alarm distribution coefficient, and use the adjusted alarm distribution coefficient as the output result of the device alarm module; The state detection module is used to identify the working state of the IoT device when it is connected, identify the workload corresponding to the current working state of the IoT device, analyze the workload of the IoT device and the number of connected devices in each cycle, and obtain the state recognition coefficient corresponding to the working state; The quality verification module is used to determine the working model of the currently connected IoT device according to the output results of the data transmission module, the device alarm module and the status detection module, and manage the connected IoT device through the obtained working model; The transmission stability coefficient is obtained by obtaining the data packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability of the IoT device during data transmission, and then the data packet loss rate, average delay, jitter duration, retransmission rate, number of connection interruptions and availability are standardized to obtain the transmission stability coefficient. ； in, Expressed as the transmission stability coefficient, represents the packet loss rate, represents the average delay, Indicates the jitter duration. Represents the retransmission rate, Indicates the number of connection interruptions. Indicates availability; represents the weight of the packet loss rate, represents the weight of the average delay, Indicates the weight of the jitter duration, represents the weight of the retransmission rate, Indicates the weight of the number of connection interruptions, The weight representing the availability; The method for obtaining the alarm distribution coefficient is as follows: selecting the measurement points whose values ​​exceed the safety range, and treating the measurement points that exceed the safety range as alarm points; obtaining an alarm point set S, where there are n alarm points in the alarm point set, calculating the alarm probability of each alarm point in the alarm point set and the joint probability of every two alarm points, and obtaining the alarm distribution coefficient according to the calculated alarm probability and joint probability; ； in, Represents the alarm distribution coefficient of the alarm point set, Indicates the number of alarm points in the alarm point set. represents the weight of the i-th alarm point, represents the alarm probability of the ith alarm point, represents the adjustment parameter; Indicates the alarm point set. Point pairs indicating alarm points; represents the joint weight of the i-th alarm point and the j-th alarm point, represents the joint probability of the i-th alarm point and the j-th alarm point; The alarm distribution factor is adjusted as follows: Verify whether the alarm distribution coefficient calculated at this time is greater than the average value of the alarm distribution coefficient in the historical data. If it is greater than the average value of the alarm distribution coefficient in the historical data, sort the alarm points from large to small according to the values ​​of the alarm probability and the joint probability, and select the alarm points whose alarm probability and joint probability are both in the top 10% and the bottom 10% to form an alarm point adjustment set; calculate the alarm distribution coefficient of the alarm point adjustment set, adjust the weight and joint weight of the alarm point according to the alarm distribution coefficient of the alarm point adjustment set and the alarm distribution coefficient of the alarm point set, adjust the alarm distribution coefficient according to the adjusted alarm point weight and joint weight, and use the adjusted alarm distribution coefficient as the alarm distribution coefficient output by the equipment alarm module; The implementation method of adjusting the weight of the alarm point is as follows: ； in, represents the weight of the i-th alarm point after adjustment, represents the weight of the i-th alarm point, Represents the alarm distribution coefficient of the alarm point set, Indicates the alarm distribution coefficient of the alarm point adjustment set, Represents the average value of the alarm distribution coefficient in historical data, Indicates the number of alarm points in the alarm point set. Indicates the number of alarm points in the alarm point adjustment set. represents the exponential constant; The implementation of adjusting the joint weight is as follows: ； in, represents the joint weight of the adjusted i-th alarm point and the j-th alarm point, represents the joint weight of the i-th alarm point and the j-th alarm point, represents the exponential constant, Represents the alarm distribution coefficient of the alarm point set, Indicates the alarm distribution coefficient of the alarm point adjustment set, Indicates the number of alarm points in the alarm point set. Indicates the number of alarm points in the alarm point adjustment set; The state recognition coefficient is expressed as follows: ； in, represents the state recognition coefficient, represents the weight of the workload, Expressed as the transmission stability coefficient, Represents the workload.

2. A unified access management system for traffic Internet of Things sensing devices according to claim 1, characterized in that: The implementation method of the workload corresponding to the current working state of the IoT device is to obtain the CPU usage rate, memory usage rate and network bandwidth usage rate of the IoT device, and use the weighted sum of the CPU usage rate, memory usage rate and network bandwidth usage rate as the workload corresponding to the current working state of the IoT device; The workload is represented as follows: ； in, Expressed as workload, Expressed as CPU usage, Expressed as memory usage, Expressed as network bandwidth usage, Expressed as the weight of CPU usage, Expressed as the weight of memory usage, Expressed as the weight of network bandwidth usage.

3. A unified access management system for traffic Internet of Things sensing devices according to claim 1, characterized in that: The implementation method of analyzing the workload of IoT devices and the number of connected devices in each cycle is to establish a relationship model between the workload and the number of connected devices through statistical analysis of historical data; The acquired relationship model is verified to determine the determination coefficient of the relationship model. When the determination coefficient is greater than the standard determination value, the state recognition coefficient is obtained according to the currently calculated workload and the transmission stability coefficient.

4. A unified access management system for traffic Internet of Things sensing devices according to claim 1, characterized in that: The implementation of the working model is as follows: According to the output results of the data transmission module, the interface alarm module and the status detection module, the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment are extracted; the working status, status recognition coefficient, alarm distribution coefficient and transmission stability coefficient of the equipment are combined into a working model, and the boundary value of the working model is determined; The boundary value of the working model is used as the retrieval condition, the boundary conditions corresponding to the boundary value are traversed, and the boundary condition with the greatest similarity among the boundary conditions is used as the output working model.

Images