CN-122020267-A - Tray dynamic tracking method and device based on Internet of things

CN122020267ACN 122020267 ACN122020267 ACN 122020267ACN-122020267-A

Abstract

The embodiment of the application provides a method and a device for dynamically tracking a tray based on the Internet of things, which realize accurate acquisition of positioning through signal fusion and confidence calculation. And constructing a tracking mechanism, combining state classification and window regulation, and establishing a reliable track generation strategy. Track optimization is introduced, and continuous improvement of tracking is ensured through breakpoint detection and probability filling. The method effectively solves the defects of the traditional technology in the aspects of data acquisition, state identification, track processing and the like, and provides technical support for tray tracking.

Inventors

XU ZHIHAO
WANG KAI

Assignees

浙江久鼎智联科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The method for dynamically tracking the tray based on the Internet of things is characterized by comprising the following steps: the edge node acquires tray sensing data through a radio frequency identification reader-writer and a Bluetooth positioning base station and gathers the tray sensing data according to a preset message queue protocol to obtain an original sensing data set, confidence calculation is carried out on the original sensing data set according to a signal intensity threshold condition and a historical stability measure to obtain confidence sensing data, and weighting and fusion are carried out on the confidence sensing data with the same tray identification according to a time stamp alignment rule to obtain a space-time feature vector; Performing state classification according to a preset motion judgment threshold condition based on the position difference value sequence and the direction change amplitude of the space-time feature vector to obtain a motion state mark, determining the length of a sampling window according to a preset window regulation rule according to the motion state mark to obtain sampling window configuration, and aggregating the space-time feature vector in the sampling window configuration range according to a weighted average rule to obtain a perception event; The cloud receives the sensing event and arranges the sensing event according to the tray identification and the time stamp to obtain an original track sequence, the original track sequence is subjected to regional mapping according to the preset semantic region boundary definition to obtain a semantic track sequence, the semantic track sequence is detected according to the breakpoint judgment threshold condition to obtain a track breakpoint set, probability filling is carried out according to a track inference model based on the track breakpoint set to obtain a complete track record, and the complete track record is written into persistent storage to generate an alarm event.
2. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein the edge node obtains tray sensing data through the radio frequency identification reader and the bluetooth positioning base station and gathers the tray sensing data according to a preset message queue protocol to obtain an original sensing data set, and the method comprises the following steps: The edge node acquires tray identification and signal intensity data through a radio frequency identification reader-writer, acquires tray position coordinates and timestamp data through a Bluetooth positioning base station to generate a multi-source acquisition data stream, and performs field extraction and format standardization processing on the multi-source acquisition data stream according to a preset data analysis rule to generate a standardized perception data sequence; and performing channel access and buffer queue writing on the standardized sensing data sequence according to a preset message queue protocol to generate a data queue to be aggregated, performing grouping aggregation on the data queue to be aggregated according to a tray identifier, and performing ascending arrangement according to a time stamp to generate an original sensing data set.
3. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein the performing confidence calculation on the original sensing data set according to a signal strength threshold condition and a historical stability metric to obtain confidence sensing data, and performing weighted fusion on the confidence sensing data identified by the same tray according to a timestamp alignment rule to obtain a space-time feature vector comprises: Comparing and judging the signal intensity of each piece of data in the original perception data set according to a preset signal intensity threshold condition to generate a signal quality score, extracting a historical perception data sliding window according to a tray mark, calculating the space consistency deviation of the current data and the data in the historical perception data sliding window to generate a stability metric value, carrying out weighted summation on the signal quality score and the stability metric value according to a preset weight configuration to generate a confidence evaluation value, and packaging the confidence evaluation value and the corresponding data to generate confidence perception data; And carrying out time stamp calibration on the sensing data with the confidence coefficient identified by the same tray according to the standard time acquired by the network time protocol to generate a time alignment sensing data sequence, and carrying out weighted fusion on position coordinates and motion directions on a plurality of pieces of data with similar time stamps in the time alignment sensing data sequence according to fusion weights normalized by respective confidence coefficient evaluation values to generate a space-time feature vector.
4. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein the step of classifying the states of the position difference sequence and the direction change amplitude based on the space-time feature vector according to a preset movement judgment threshold condition to obtain movement state marks comprises the steps of: Calculating the position coordinate difference value of adjacent vectors according to the time stamp sequence to generate a position difference value sequence, and calculating the moving direction included angle of the adjacent vectors according to the time stamp sequence to generate a direction change amplitude sequence; And carrying out low displacement judgment in a continuous period on the position difference value sequence according to a preset stillness judgment threshold condition, carrying out direction mutation judgment on the direction change amplitude sequence according to a preset jerkiness threshold condition to generate a movement characteristic judgment result, and carrying out category classification of stillness state, steady movement state and rapid change state on the movement characteristic judgment result according to a preset state classification rule to generate a movement state mark.
5. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein determining the sampling window length according to the motion state mark and the preset window adjustment rule to obtain the sampling window configuration, and aggregating the space-time feature vectors in the sampling window configuration range according to the weighted average rule to obtain the sensing event comprises: Matching and inquiring the motion state mark with a state window mapping table in a preset window regulation rule to obtain window length parameters of corresponding states, and calculating window start-stop time of the window length parameters and a current time stamp to generate sampling window configuration; And carrying out weighted average calculation on the position coordinates of the space-time feature vectors in the sampling window configuration range according to the weight normalized by the confidence coefficient evaluation value of each vector to generate event positions, and packaging the event positions, the start and stop time stamps configured in the sampling window, the motion state marks and the average confidence coefficient in the window to generate a perception event.
6. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein the cloud terminal receives the sensing event and arranges the sensing event according to the tray identifier and the timestamp to obtain an original track sequence, and performs region mapping on the original track sequence according to a preset semantic region boundary definition to obtain a semantic track sequence, and the method comprises the following steps: The cloud acquires the sensing events uploaded by each edge node through a message receiving interface, groups the sensing events according to a tray identifier to generate a tray event grouping set, and arranges the sensing events in each group in the tray event grouping set according to the ascending order of time stamps to generate an original track sequence; And carrying out coordinate attribution judgment on event positions of each sensing event in the original track sequence according to preset semantic region boundary definition to generate a region attribution result, converting the region attribution result into semantic region codes, adding region hierarchy labels, and carrying out field expansion packaging with the corresponding sensing event to generate a semantic track sequence.
7. The method for dynamically tracking the tray based on the internet of things according to claim 1, wherein the detecting the semantic track sequence according to a breakpoint judgment threshold condition to obtain a track breakpoint set, probability filling according to a track inference model based on the track breakpoint set to obtain a complete track record, writing the complete track record into persistent storage and generating an alarm event comprises: performing interval calculation on time stamps of adjacent perceived events in the semantic track sequence to generate a time interval sequence, performing overrun detection on the time interval sequence according to a preset breakpoint judgment threshold condition, and extracting semantic areas of front and rear events corresponding to overrun positions and the time stamps to generate a track breakpoint set; And calculating candidate path probability of each breakpoint in the track breakpoint set according to space-time feature vectors of events before and after the breakpoint and a historical circulation path statistics call track inference model, selecting a path with highest posterior probability to generate a probability filling interval, inserting the probability filling interval into a corresponding breakpoint position of the semantical track sequence to generate a complete track record, writing the complete track record into persistent storage, and carrying out anomaly detection on the complete track record according to a preset stay threshold condition and a standard circulation path rule to generate an alarm event.
8. A tray dynamic tracking device based on the internet of things, the device comprising: The tray sensing module is used for acquiring tray sensing data by the edge node through the radio frequency identification reader-writer and the Bluetooth positioning base station, converging the tray sensing data according to a preset message queue protocol to obtain an original sensing data set, performing confidence calculation on the original sensing data set according to a signal strength threshold condition and a historical stability measure to obtain confidence sensing data, and performing weighted fusion on the confidence sensing data with the same tray mark according to a timestamp alignment rule to obtain a space-time feature vector; The motion marking module is used for carrying out state classification according to a preset motion judgment threshold condition based on the position difference value sequence and the direction change amplitude of the space-time feature vector to obtain a motion state mark, determining the length of a sampling window according to a preset window regulation rule according to the motion state mark to obtain sampling window configuration, and carrying out aggregation on the space-time feature vector in the sampling window configuration range according to a weighted average rule to obtain a perception event; The track tracking module is used for receiving the perception event by the cloud end, arranging the perception event according to the tray identification and the time stamp to obtain an original track sequence, carrying out region mapping on the original track sequence according to the preset semantic region boundary definition to obtain a semantic track sequence, detecting the semantic track sequence according to the breakpoint judgment threshold condition to obtain a track breakpoint set, carrying out probability filling on the track breakpoint set according to a track inference model to obtain a complete track record, writing the complete track record into persistent storage, and generating an alarm event.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the internet of things based tray dynamic tracking method of any one of claims 1 to 7 when the program is executed by the processor.
10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the internet of things based tray dynamic tracking method of any of claims 1 to 7.

Description

Tray dynamic tracking method and device based on Internet of things Technical Field The application relates to the field of data processing, in particular to a method and a device for dynamically tracking a tray based on the Internet of things. Background The existing tray tracking method has obvious defects. The traditional system has poor performance in terms of data acquisition and signal processing, and cannot effectively realize accurate positioning of the tray, so that the tracking effect is affected. Furthermore, the prior art has bottlenecks in state recognition and trajectory generation. Most systems lack perfect motion decision mechanisms and sampling adjustment strategies, resulting in less than ideal track recording accuracy. Existing systems have technology shortboards in terms of track handling. The lack of deep analysis of break points makes it difficult to achieve efficient trajectory completion through probabilistic filling, affecting tracking accuracy. The resolution of these problems is of great importance for improving the tracking ability of the pallet. Disclosure of Invention Aiming at the problems in the prior art, the application provides a tray dynamic tracking method and device based on the Internet of things, which can effectively solve the defects of the traditional technology in the aspects of data acquisition, state identification, track processing and the like and provide technical support for tray tracking. In order to solve at least one of the problems, the application provides the following technical scheme: in a first aspect, the present application provides a method for dynamically tracking a tray based on the internet of things, including: the edge node acquires tray sensing data through a radio frequency identification reader-writer and a Bluetooth positioning base station and gathers the tray sensing data according to a preset message queue protocol to obtain an original sensing data set, confidence calculation is carried out on the original sensing data set according to a signal intensity threshold condition and a historical stability measure to obtain confidence sensing data, and weighting and fusion are carried out on the confidence sensing data with the same tray identification according to a time stamp alignment rule to obtain a space-time feature vector; Performing state classification according to a preset motion judgment threshold condition based on the position difference value sequence and the direction change amplitude of the space-time feature vector to obtain a motion state mark, determining the length of a sampling window according to a preset window regulation rule according to the motion state mark to obtain sampling window configuration, and aggregating the space-time feature vector in the sampling window configuration range according to a weighted average rule to obtain a perception event; The cloud receives the sensing event and arranges the sensing event according to the tray identification and the time stamp to obtain an original track sequence, the original track sequence is subjected to regional mapping according to the preset semantic region boundary definition to obtain a semantic track sequence, the semantic track sequence is detected according to the breakpoint judgment threshold condition to obtain a track breakpoint set, probability filling is carried out according to a track inference model based on the track breakpoint set to obtain a complete track record, and the complete track record is written into persistent storage to generate an alarm event. The edge node obtains the tray identification and signal intensity data through the radio frequency identification reader-writer, obtains the tray position coordinates and the time stamp data through the Bluetooth positioning base station to generate a multi-source acquisition data stream, and performs field extraction and format standardization processing on the multi-source acquisition data stream according to a preset data analysis rule to generate a standardized perception data sequence; and performing channel access and buffer queue writing on the standardized sensing data sequence according to a preset message queue protocol to generate a data queue to be aggregated, performing grouping aggregation on the data queue to be aggregated according to a tray identifier, and performing ascending arrangement according to a time stamp to generate an original sensing data set. Further, the method further comprises the steps of comparing the signal intensity of each piece of data in the original perception data set according to a preset signal intensity threshold condition to generate a signal quality score, extracting a historical perception data sliding window according to tray identification for each piece of data, calculating the space consistency deviation of the current data and the data in the historical perception data sliding window to generate a stability metric value, carrying out weighte