Remote asset management system based on passive internet of things

Abstract

The invention discloses a remote asset management system based on a passive internet of things, which is characterized in that an RFID reader-writer, a small-size passive plate-shaped antenna, wireless information acquisition equipment, FSU access equipment and the like are arranged in a communication machine room and a data machine room based on passive RFID tags, and meanwhile, a tag identification scheduling algorithm based on priority is adopted on passive RFID tag identification; the system is in butt joint with a machine room equipment asset management system to construct an intelligent management system covering the whole life cycle of machine room equipment from goods acceptance check, machine (warehouse) access, maintenance scrapping, spare part management and the like, equipment assets in a machine room can be checked remotely in real time at low cost according to needs, the manual operation and maintenance cost is greatly reduced, the equipment resource utilization rate is improved, and the working efficiency is improved. And the intelligent management level of the equipment in the machine room is improved.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology, and is particularly applicable to remote asset management systems based on passive IoT. Background Technology

[0002] Traditional communication and data centers primarily rely on manual on-site inventory checks and simple resource management systems using labels, QR codes, or passive RFID tags. This management model suffers from low inventory efficiency and unclear resource status. It cannot track equipment entry and exit from the room in real time, nor can it promptly detect changes in equipment location or dynamically monitor rack capacity. This results in high manual maintenance costs, low resource utilization, and inefficient asset inventory checks. Furthermore, it is difficult to effectively integrate with asset management systems, posing a security risk of asset loss.

[0003] Currently, in data center equipment management scenarios using passive RFID tags, besides the mainstream method of manually inventorying equipment using barcode scanners, active phased array antennas installed in the data center can also be used to automatically identify passive RFID tags and achieve automatic equipment inventory. However, active phased array antennas have problems such as large size, high power consumption, stringent requirements for the data center environment, and high false negative rate, which prevents their large-scale promotion.

[0004] This invention addresses the current situation of low efficiency and error-prone management of communication equipment rooms and data centers, the inability to timely grasp resource usage, and the resulting low resource utilization rate. It proposes a remote asset management system based on passive Internet of Things. Summary of the Invention

[0005] The purpose of this invention is to provide a remote asset management system based on passive Internet of Things (IoT) to solve problems such as low management efficiency and susceptibility to errors in communication equipment rooms and data centers, the inability to keep track of resource usage in a timely manner, and low resource utilization.

[0006] To achieve the above objectives, the remote asset management system based on passive Internet of Things (IoT) of the present invention includes passive RFID tags, RFID tag identification antennas, RFID readers, and on-site environmental monitoring units. The RFID reader has a built-in priority-based tag identification scheduling algorithm that dynamically assigns a priority to each passive RFID tag based on its real-time status and business attributes, and then allocates response time slots based on the priority to orderly acquire passive RFID tag information. Furthermore, it utilizes the existing environmental monitoring units in the computer room to connect to a remote equipment asset management center, thereby realizing automated, real-time, and full-process management of computer room equipment assets.

[0007] Furthermore, the RFID tag identification antenna is a passive high-gain circularly polarized plate antenna.

[0008] Furthermore, the priority-based tag identification scheduling algorithm specifically includes parameter initialization, real-time acquisition of RFID tag status, priority calculation, response time slot allocation, and algorithm iterative optimization. Parameter initialization is initiated immediately after the RFID reader / writer starts to complete the basic parameter configuration of the algorithm. Real-time acquisition of RFID tag status is used to dynamically obtain the status data of passive RFID tags. Priority calculation is used to determine the priority of each passive RFID tag. Response time slot allocation allocates response time slots to each passive RFID tag according to priority to obtain passive RFID tag information in an orderly manner. Algorithm iterative optimization optimizes the basic parameters of the algorithm based on the identification results of the passive RFID tags.

[0009] Furthermore, the parameter initialization specifically includes the following steps: S1.1 Initialize the basic parameters of the algorithm according to the scenario requirements and business rules; including the signal strength benchmark threshold of passive RFID tags, priority level settings, resource ratios corresponding to each priority level, and weights of business attributes; S1.2, the RFID reader broadcasts an initialization command to complete a handshake with the passive RFID tag; S1.3 Record the RFID tag ID and initial signal strength value based on the basic information fed back by the passive RFID tag.

[0010] Furthermore, the business attributes include equipment value, signal strength, and timeliness; the weights of the business attributes are equipment value weight, signal strength weight, and timeliness weight.

[0011] Furthermore, the real-time acquisition of RFID tag status specifically includes the following steps: S2.1, the RFID reader sends status query commands to the passive RFID tag at preset intervals; S2.2, the RFID reader analyzes the real-time signal strength of the passive RFID tag's backscattered signal through the signal processing module and records the RSSI value of each passive RFID tag. S2.3, the RFID reader uses multiple antennas for positioning and calculates the relative position of the passive RFID tag using the angle of arrival, marking the passive RFID tag located in the signal-blocked area; S2.4 Record the ID, RSSI value, location, and service attributes of each passive RFID tag.

[0012] Furthermore, the RSSI value of each passive RFID tag is determined by the RFID reader continuously collecting the RSSI value of the same passive RFID tag three times, and then eliminating environmental noise interference through moving average filtering.

[0013] Furthermore, the response time slot allocation specifically includes the following steps: S3.1 Calculate the scores of each business attribute for each passive RFID tag; S3.2, calculate the comprehensive priority score of each passive RFID tag by weighted summation; S3.3, determine the priority of each passive RFID tag based on the comprehensive priority score.

[0014] Furthermore, the priority calculation specifically includes the following steps: S4.1, set the basic time unit of the scheduling frame, and allocate a scheduling frame response time slot to each priority according to the resource allocation ratio corresponding to each priority; S4.2, Based on the priority of each passive RFID tag, a hash mapping is used to allocate a dedicated response time slot for all passive RFID tags under each priority. S4.3, the RFID reader broadcasts time slot instructions in time slot sequence to orderly acquire passive RFID tag information.

[0015] Furthermore, the algorithm iterative optimization specifically includes the following steps: S5.1, Calculate the recognition rate for each priority level; S5.2, for priority classes that fail to meet the recognition rate target, adjust the algorithm parameters, reconfigure the algorithm parameters, and execute them in the next scheduling frame.

[0016] The advantages of this invention lie in its use of passive RFID tags in communication and data centers. It involves installing RFID readers, small-sized passive panel antennas, wireless information acquisition equipment, and FSU access devices within the data center. Furthermore, it employs a priority-based tag identification scheduling algorithm for passive RFID tag recognition, interfacing with the data center equipment asset management system. This constructs an intelligent management system covering the entire lifecycle of data center equipment, from arrival and acceptance, entry and exit from the data center, maintenance and disposal, and spare parts management. This system enables low-cost, on-demand, real-time, and remote inventory management of equipment assets within the data center, significantly reducing manual maintenance costs, improving equipment resource utilization, and ultimately enhancing the intelligent management level of data center equipment. Attached Figure Description

[0017] Figure 1 This invention relates to a remote asset management system based on passive Internet of Things (IoT).

[0018] Figure 2 This invention provides a flowchart for parameter initialization of a remote asset management system based on passive Internet of Things (IoT).

[0019] Figure 3 This invention presents a flowchart of the real-time data acquisition process for RFID tag status in a remote asset management system based on passive Internet of Things (IoT).

[0020] Figure 4 This invention presents a flowchart of priority calculation for a remote asset management system based on passive Internet of Things.

[0021] Figure 5 This invention presents a flowchart of the response time slot allocation process for a remote asset management system based on passive Internet of Things.

[0022] Figure 6 This invention presents a flowchart of the algorithm iteration optimization process for a remote asset management system based on passive Internet of Things. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] The remote asset management system based on passive Internet of Things described in this invention, such as... Figure 1 As shown, the system includes passive RFID tags, RFID tag identification antennas, RFID readers, and on-site environmental monitoring units. Passive RFID tags are directly affixed to or installed on the equipment requiring management. The RFID tag identification antenna establishes a contactless energy transfer and data communication bridge between the passive RFID tag and the RFID reader via radio frequency signals. When the RFID tag identification antenna receives electromagnetic waves of a specific frequency emitted by the RFID reader, it converts them into weak electrical energy, activating the passive RFID tag chip and simultaneously modulating and reflecting the data stored within the passive RFID tag chip (such as a unique identification code) back to the RFID reader.

[0025] The RFID reader has a built-in priority-based tag identification and scheduling algorithm. Based on the real-time status of the passive RFID tag and business needs, it dynamically assigns a priority to each passive RFID tag and then allocates response time slots based on the priority to obtain passive RFID tag information in an orderly manner.

[0026] Then, by utilizing the existing environmental monitoring units in the computer room, we can connect to the remote equipment asset management center to achieve automated, real-time, and full-process management of the computer room equipment assets.

[0027] The RFID tag identification antenna adopts a passive high-gain circularly polarized plate antenna, which has the advantages of small size, low power consumption, easy installation and low cost compared with active phased array antennas. It is conducive to low-cost and large-scale deployment, and can be flexibly adapted to different data center environments by increasing or decreasing the number of antennas, ultimately achieving an identification success rate of over 99%.

[0028] The RFID reader has a built-in priority-based tag identification scheduling algorithm. During the identification process, it solves three core problems: "weak signal tags being masked", "local dense tag collisions", and "high-value tag response delay" through differentiated resource allocation.

[0029] Specifically, the priority-based tag identification scheduling algorithm assigns a dynamic priority to each passive RFID tag based on its real-time status (signal quality, location) and business attributes (equipment value, timeliness). Then, it schedules response slots based on these priorities, ultimately achieving the effect of "no weak-signal RFID tags being missed, high-value RFID tags being prioritized for identification, and overall collision rate reduction." For example, tags with weak signals at the bottom or corners of the data center are given priority to avoid missed reads; high-value equipment tags (such as switches and servers) are assigned high priority to ensure real-time tracking; and mobile equipment tags (which require real-time monitoring of their specific locations) are assigned higher priority than inventory equipment tags to ensure the controllability and manageability of mobile equipment.

[0030] The priority-based tag identification scheduling algorithm specifically includes parameter initialization, real-time acquisition of RFID tag status, priority calculation, response time slot allocation, and algorithm iterative optimization.

[0031] After the RFID reader / writer starts up, it immediately enters the algorithm parameter initialization stage. This initialization completes the configuration of the algorithm's basic parameters, providing baseline rules for subsequent algorithms and preventing scheduling chaos due to missing parameters. For example... Figure 2 As shown, the specific steps include: S1.1 Initialize the basic algorithm parameters according to scenario and business requirements. This includes: Set a baseline threshold for the signal strength of passive RFID tags, i.e., the RSSI threshold. For example, -60dBm is a strong signal, -85dBm is a weak signal threshold, and anything below -95dBm is considered an extremely weak signal.

[0032] Set priority levels. Specifically, set the priority to five levels, from P0 to P4, with P0 being the highest and P4 the lowest.

[0033] Set the resource ratio corresponding to each priority level, such as P0 priority using 25% of resources, P1 priority using 20% ​​of resources, P2 priority using 15% of resources, P3 priority using 15% of resources, and P4 priority using 25% of resources.

[0034] Set weight coefficients for each business attribute. Business attributes include equipment value, signal strength, and timeliness. The weights for business attributes are: equipment value weight, signal strength weight, and timeliness weight. For example, assign a weight of W1=0.4 to equipment value, W2=0.3 to signal strength, and W3=0.3 to timeliness to ensure that priority calculations align with the requirements of the business attributes.

[0035] S1.2, the RFID reader broadcasts an initialization command to complete a handshake with the passive RFID tags. The initialization command broadcast by the RFID reader triggers the wake-up of all passive tags within the coverage area and obtains basic information from the passive RFID tags, such as RFID tag ID and initial signal strength.

[0036] S1.3 Based on the basic information fed back by the passive RFID tags, the RFID reader establishes an initial tag list, records the RFID tag ID, initial RSSI value, etc., and excludes unresponsive and invalid passive RFID tags.

[0037] Real-time RFID tag status acquisition is used to dynamically obtain the status data of passive RFID tags, providing accurate input data for subsequent priority calculations in the algorithm and avoiding time slot allocation based on outdated passive RFID tag status. For example... Figure 3 As shown, the specific steps include: S2.1, the RFID reader sends status query commands to the passive RFID tags according to the initial tag list at preset intervals. The preset interval can be adjusted according to different scenarios, such as 50ms for data center scenarios, 200ms for communication room scenarios, and 100ms for general scenarios.

[0038] S2.2, the RFID reader analyzes the real-time signal strength (RSSI) of the passive RFID tag's backscattered signal through its signal processing module. The RSSI value of each passive RFID tag is recorded. For example, the real-time RSSI of passive RFID tag ID=001 is -78dBm. The RSSI value of each passive RFID tag is determined by the RFID reader continuously collecting the RSSI value of the same passive RFID tag three times, and then filtering it using a moving average to eliminate environmental noise interference.

[0039] S2.3, the RFID reader uses multiple antennas for positioning and calculates the relative position of passive RFID tags using the angle of arrival (AoA), marking passive RFID tags located in areas with signal obstruction. Simultaneously, it reads the built-in business attributes of the passive RFID tags, such as associating item information through device coding; identifying high-value tags, such as tags with the ID prefix "VIP-"; and time-sensitive tags, such as tags with the ID prefix "Flow-".

[0040] In S2.4, the RFID reader records and organizes the read passive RFID tag ID, RSSI value, location, service attributes, etc., to form a real-time tag status table.

[0041] Based on the real-time status of the passive RFID tag (obtained from the real-time tag status table in step S2.4), combined with the weights of the service attributes in the algorithm initialization parameter table, the priority of each tag is calculated using a quantization model, providing a basis for subsequent response time slot allocation. As Figure 4 shown, it specifically includes the following steps: In S3.1, calculate the service scores of each passive RFID tag respectively.

[0042] For the signal strength, according to the preset mapping relationship between the RSSI value and the signal strength score, determine the signal strength score of each passive RFID tag.

[0043] When RSSI ≥ -60 dBm, the signal strength score S2 = 10 points. This is a strong signal at this time.

[0044] When -85 dBm < RSSI < -60 dBm, the signal strength score S2 = (RSSI + 95) / 3.5 points.

[0045] If RSSI is -78 dBm, then the signal strength score S2 = (-78 + 95) / 3.5 ≈ 4.86 points.

[0046] When RSSI ≤ -85 dBm, the signal strength score S2 = 0 points. This is a weak signal at this time.

[0047] For the equipment value, according to the preset standard, high-value equipment (such as unit price ≥ 1000 yuan) is given an equipment value score S1 = 10 points; medium-value equipment (such as 100 yuan ≤ unit price < 1000 yuan) is given an equipment value score S1 = 6 points; low-value equipment (such as unit price < 100 yuan) is given an equipment value score S1 = 2 points.

[0048] For the timeliness, based on the expiration time node, equipment with remaining time ≤ 24 hours is given a timeliness score S3 = 10 points; equipment with 24 hours < remaining time ≤ 72 hours is given a timeliness score S3 = 6 points; equipment with remaining time > 72 hours is given a timeliness score S3 = 2 points.

[0049] In S3.2, calculate the comprehensive priority score of each passive RFID tag.

[0050] Based on the weights of business attributes, such as assigning a weight of 0.4 to device value, 0.3 to signal strength, and 0.3 to timeliness, a weighted summation method is used to calculate the overall priority score P_score for each passive RFID tag. The formula is: P_score = W1×S1 + W2×(10 - S2) + W3×S3 Among them, "10 - S2" means that the weaker the signal, the higher the contribution score of the signal strength to the overall priority, which is consistent with the priority requirements.

[0051] For example, for a high-value device tag (S1=10, S3=10), the real-time RSSI is -88dBm (S2=0), and the weights are W1=0.4, W2=0.3, and W3=0.3, then: P_score = 0.4×10 + 0.3×(10-0) + 0.3×10 = 4 + 3 + 3 = 10 points.

[0052] S3.3 Determine the priority of each passive RFID tag based on the comprehensive priority score. Specifically, the comprehensive priority score P_score is divided into 5 levels, corresponding to 5 priorities. P_score ≥ 9 corresponds to the highest priority P0; 7 ≤ P_score < 9 corresponds to priority P1; 5 ≤ P_score < 7 corresponds to priority P2; 3 ≤ P_score < 5 corresponds to priority P3; and P_score < 37 corresponds to the lowest priority P4.

[0053] The goal of response slot allocation is to translate priorities into actual response slots, ensuring that high-priority passive RFID tags get more opportunities for collision-free detection, while also preventing low-priority passive RFID tags from being missed. For example... Figure 5 As shown, the specific steps include: S4.1 sets the basic time unit for scheduling frames and allocates scheduling frame response slots to each priority level according to the resource allocation ratio corresponding to each priority level. For example, if the basic time unit for scheduling frames is set to 1 second / frame, each scheduling frame is divided into 100 equal slots. If 1 second / frame is used, then each slot is 10ms. Based on the resource ratio corresponding to each priority level, such as P0 priority occupying 25% of the resources, P1 priority occupying 20% ​​of the resources, P2 priority occupying 15% of the resources, P3 priority occupying 15% of the resources, and P4 priority occupying 25% of the resources, then the corresponding priority levels are: P0 priority occupying 25 scheduling frame slots, P1 priority occupying 20 scheduling frame slots, P2 priority occupying 15 scheduling frame slots, P3 priority occupying 15 scheduling frame slots, and P4 priority occupying 25 scheduling frame slots.

[0054] A scheduling frame time slot allocation table is generated for each priority level. For example, priority P0 occupies 25 scheduling frame time slots, with time slot numbers 1-25; priority P1 occupies 20 scheduling frame time slots, with time slot numbers 26-45; priority P2 occupies 15 scheduling frame time slots, with time slot numbers 46-60; priority P3 occupies 15 scheduling frame time slots, with time slot numbers 61-75; and priority P4 occupies 25 scheduling frame time slots, with time slot numbers 76-100.

[0055] S4.2, based on the priority of each passive RFID tag, a hash mapping is used to allocate a dedicated time slot for all passive RFID tags under each priority. Specifically, after converting the passive RFID tag ID to a numerical value, the modulo of the number of scheduling frame time slots corresponding to its priority is taken to obtain the fixed time slot number for that passive RFID tag ID. For example, if the passive RFID tag ID = 001, the ID is converted to the numerical value 1. The priority of this passive RFID tag is P0. Priority P0 occupies 20 scheduling frame time slots, meaning the number of scheduling frame time slots is 20. 1 mod 25 = 1, therefore the dedicated time slot number corresponding to passive RFID tag ID = 001 is 1, thus avoiding collisions between tags of the same priority in the same time slot.

[0056] S4.3, the RFID reader broadcasts time slot instructions sequentially to acquire passive RFID tag information in an orderly manner. The time slot instruction contains the priority level and tag ID corresponding to the current time slot, such as "Time Slot 1, P0, ID=001". Only passive RFID tags matching this time slot instruction (ID=001) send response data, which is received and parsed by the RFID reader. If no tag responds in a time slot, meaning the time slot is idle, the RFID reader automatically allocates the time slot temporarily to the next priority level. For example, if time slot 2 with priority P0 is idle, the idle time slot is allocated to tags with priority P1, improving resource utilization.

[0057] The algorithm iteratively optimizes the identification results of passive RFID tags, improving the priority calculation rules and time slot allocation ratio to avoid the problem of "static rules not adapting to dynamic scenarios." For example... Figure 6 As shown, the specific steps include: S5.1, Calculate the recognition rate for each priority level: Recognition rate = (Number of successfully recognized tags / Total number of tags for that priority level) × 100%.

[0058] S5.2 adjusts the algorithm parameters for priority classes that fail to meet the recognition rate target and executes them in the next scheduling frame.

[0059] If the P0 recognition rate is less than 80%, the algorithm parameter table will be adjusted and the algorithm parameters will be reconfigured. This includes increasing time slot quotas, refining time slot allocation, and adjusting business weights.

[0060] For example, if the P0 priority recognition rate is low and does not meet the standard, the time slot quota of P0 is increased, such as from 25% to 30%, or the signal strength weight W2 is increased from 0.3 to 0.4.

[0061] If the recognition rate of P4 priority is low and fails to meet the standard due to frequent collisions, the time slot allocation of P4 should be refined, such as splitting the 25 time slots into smaller granularities to reduce collisions of the same priority.

[0062] If business needs change, business weights can be temporarily adjusted, such as increasing the timeliness weight W3 for mobile devices from 0.3 to 0.5.

Claims

1. A remote asset management system based on passive Internet of Things (IoT), comprising passive RFID tags, RFID tag identification antennas, RFID readers, and on-site environmental monitoring units; characterized in that: The RFID reader has a built-in priority-based tag identification and scheduling algorithm. Based on the real-time status and business attributes of the passive RFID tags, it dynamically assigns a priority to each passive RFID tag and then allocates response time slots based on the priority to obtain passive RFID tag information in an orderly manner. It can also connect to the remote equipment asset management center by utilizing the existing environmental monitoring unit in the computer room to achieve automated, real-time, and full-process management of the computer room equipment assets.

2. The remote asset management system based on passive Internet of Things as described in claim 1, characterized in that: The RFID tag identification antenna is a passive high-gain circularly polarized plate antenna.

3. The remote asset management system based on passive Internet of Things as described in claim 1, characterized in that: The priority-based tag identification scheduling algorithm specifically includes parameter initialization, real-time acquisition of RFID tag status, priority calculation, response time slot allocation, and algorithm iterative optimization. Parameter initialization is initiated immediately after the RFID reader starts to complete the basic parameter configuration of the algorithm. Real-time acquisition of RFID tag status is used to dynamically obtain the status data of passive RFID tags. Priority calculation is used to determine the priority of each passive RFID tag. Response time slot allocation allocates response time slots to each passive RFID tag according to priority to obtain passive RFID tag information in an orderly manner. Algorithm iterative optimization optimizes the basic parameters of the algorithm based on the identification results of the passive RFID tags.

4. The remote asset management system based on passive Internet of Things according to claim 3, characterized in that: The parameter initialization specifically includes the following steps. S1.1 Initialize the basic parameters of the algorithm according to the scenario requirements and business rules; including the signal strength benchmark threshold of passive RFID tags, priority level settings, resource ratios corresponding to each priority level, and weights of business attributes; S1.2, the RFID reader broadcasts an initialization command to complete a handshake with the passive RFID tag; S1.3 Record the RFID tag ID and initial signal strength value based on the basic information fed back by the passive RFID tag.

5. The remote asset management system based on passive Internet of Things as described in claim 4, characterized in that: The business attributes include equipment value, signal strength, and timeliness; the weights of the business attributes are equipment value weight, signal strength weight, and timeliness weight.

6. The remote asset management system based on passive Internet of Things according to claim 3, characterized in that: The real-time acquisition of RFID tag status specifically includes the following steps. S2.1, the RFID reader sends status query commands to the passive RFID tag at preset intervals; S2.2, the RFID reader analyzes the real-time signal strength of the passive RFID tag's backscattered signal through the signal processing module and records the RSSI value of each passive RFID tag. S2.3, the RFID reader uses multiple antennas for positioning and calculates the relative position of the passive RFID tag using the angle of arrival, marking the passive RFID tag located in the signal-blocked area; S2.4 Record the ID, RSSI value, location, and service attributes of each passive RFID tag.

7. The remote asset management system based on passive Internet of Things as described in claim 6, characterized in that: The RSSI value of each passive RFID tag is determined by the RFID reader continuously collecting the RSSI value of the same passive RFID tag three times, and then filtering it through a moving average to eliminate environmental noise interference.

8. The remote asset management system based on passive Internet of Things according to claim 3, characterized in that: The response time slot allocation specifically includes the following steps. S3.1 Calculate the scores of each business attribute for each passive RFID tag; S3.2, calculate the comprehensive priority score of each passive RFID tag by weighted summation; S3.3, determine the priority of each passive RFID tag based on the comprehensive priority score.

9. The remote asset management system based on passive Internet of Things according to claim 3, characterized in that: The priority calculation specifically includes the following steps. S4.1, set the basic time unit of the scheduling frame, and allocate a scheduling frame response time slot to each priority according to the resource allocation ratio corresponding to each priority; S4.2, Based on the priority of each passive RFID tag, a hash mapping is used to allocate a dedicated response time slot for all passive RFID tags under each priority. S4.3, the RFID reader broadcasts time slot instructions in time slot sequence to orderly acquire passive RFID tag information.

10. The remote asset management system based on passive Internet of Things according to claim 3, characterized in that: The algorithm iterative optimization specifically includes the following steps. S5.1, Calculate the recognition rate for each priority level; S5.2, for priority classes that fail to meet the recognition rate target, adjust the algorithm parameters, reconfigure the algorithm parameters, and execute them in the next scheduling frame.

Images