Intelligent identification and positioning system of product packaging based on drone

Abstract

The present invention relates to the field of radio frequency identification technology, and discloses a product packaging intelligent identification and positioning system based on drones, including an intelligent identification and positioning software module and a remote monitoring module; wherein: the intelligent identification and positioning software module is used for the drone to cruise and collect ultra-high frequency tag information and the host computer to analyze and judge the collected data, and finally realize the positioning of the product packaging through the relative positioning algorithm of RSSI and phase mixing; it includes a navigation real-time processing unit, a radio frequency tag information acquisition unit, and a product packaging intelligent identification and positioning unit. The present invention can attach RFID tags to goods, and then use RFID readers to identify and track the goods, thereby realizing full-process real-time monitoring and management of the goods. The present invention uses drones equipped with RFID readers to preliminarily realize intelligent warehouse management, such as rapid positioning of goods in warehouses, inventory counting of goods, and safety of goods transportation.

Description

Technical Field

[0001] The present invention relates to the field of radio frequency identification technology, and in particular to a product packaging intelligent identification and positioning system based on an unmanned aerial vehicle. Background Art

[0002] The Internet of Things connects sensing devices through various access methods to complete information exchange, realize intelligent monitoring, control, identification, positioning, tracking, etc., covering the entire process of information collection, network transmission, data storage, data analysis and intelligent application, involving key technologies such as sensor identification, wireless communication, data storage, cloud computing, nano, and intelligent applications. Radio Frequency Identification (RFID) is one of the key technologies of the perception layer of the Internet of Things. The perception efficiency of RFID directly affects the information exchange quality of the perception layer.

[0003] RFID technology, as an important part of smart warehousing, refers to the technology that uses radio waves for identification and tracking. This technology can use readers to identify and track goods with RFID tags, thereby achieving real-time monitoring and management of the goods throughout the entire process. Some studies have proposed the use of small drones to automate the inventory process because of their autonomous mobility and the ability to hover, sense the environment and carry items.

[0004] Using drones equipped with RFID readers to realize smart warehousing can solve some problems in the process of goods management in warehouses, such as the rapid positioning of goods in warehouses, the accuracy of inventory counting, the safety of goods transportation, etc. At the same time, this technology can help companies improve the efficiency of goods transportation, reduce the intervals between consecutive scanning operations, shorten the time of goods transportation, reduce human interference factors in the warehouse management process, and ensure the safety of workers (because warehousing usually requires the ability to work on the top shelf, which puts operators at risk). Summary of the invention

[0005] In order to solve the technical problems of object identification and location tracking in smart warehousing, the present invention provides a product packaging intelligent identification and positioning system based on drones.

[0006] The present invention is implemented by the following technical solution: a product packaging intelligent identification and positioning system based on drones, including an intelligent identification and positioning software module and a remote monitoring module; wherein:

[0007] The intelligent identification and positioning software module is used for the drone to cruise and collect ultra-high frequency tag information and the host computer to analyze and judge the collected data, and finally realize the positioning of product packaging through the relative positioning algorithm of RSSI and phase mixing; it includes a navigation real-time processing unit, a radio frequency tag information acquisition unit, and a product packaging intelligent identification and positioning unit.

[0008] The remote monitoring module is used to upload the data analyzed and processed by the intelligent identification and positioning software module from the host computer to the Alibaba Cloud database, and finally display the packaging number, packaging status parameters, packaging location, warehousing records and logistics information on the Internet of Things platform to realize information sharing and remote control.

[0009] The navigation real-time processing unit checks the status of the radio frequency reader carried by the UAV before flight to ensure its normal operation, and formulates a flight plan and executes the flight mission according to the specific environment of the warehouse. During this process, feedback and adjustment of the flight status are carried out at any time.

[0010] The radio frequency tag information acquisition unit transmits information between the tag and the reader through a single-station backscatter link, thereby collecting tag information, unwrapping phase, received signal strength (RSSI), operating frequency and laser ranging information on the product packaging.

[0011] The product packaging intelligent identification and positioning unit includes packaging identity recognition, RSSI and phase mixing relative positioning algorithm.

[0012] The drone carries antenna No. 1 and antenna No. 2 that can rotate up and down. Antenna No. 1 rotates upward, and antenna No. 2 rotates downward. Antenna No. 1 and antenna No. 2 are used to collect RSSI information and phase information. Preliminary positioning information is obtained through the relative positioning algorithm mixed with RSSI and phase. At the same time, interference compensation is set to compensate and adjust the obtained preliminary information (vertical positioning information and horizontal positioning information) to obtain more accurate packaging positioning information;

[0013] Specifically, it uses the RSSI information obtained by the first antenna and the second antenna to obtain vertical positioning information through calculation, and uses the phase information obtained by the first antenna and the second antenna to obtain horizontal positioning information through calculation.

[0014] The information obtained by the first antenna includes: RSSI 1 The information obtained by antenna No. 1 and No. 2 includes: RSSI 2 and Phase II, compares the RSSI profile of the same tag from the antenna, RSSI caused by rotating the antenna 1 With RSSI 2 The difference is used to distinguish the vertical position of the shelf items, but due to the lateral polarization mismatch, it is necessary to compare the values ​​in the neighborhood of each tag RSSI peak value and define the neighborhood where γ 0 is the RSSI peak index when the antenna is located directly in front of the tag (the RSSI peak index of each tag is different, but it is a constant for a specific tag), and ε represents the neighborhood Γ εHalf the length, therefore, if we compare RSSI profiles only within the defined RSSI neighborhood, the transverse polarization mismatch will have no effect. Furthermore, taking into account the measurement noise as well as the residual errors caused by channel fading and polarization mismatch, the judgment condition can be defined as:

[0015] Level 1:

[0016] Level 2:

[0017] Level 3:

[0018] When the above residual error is taken into account, Φ is the given RSSI offset threshold.

[0019] In the above scheme, the drone is equipped with a reader and works at any height in the warehouse. It scans three shelves within the range each time and compares the RSSI difference of the dual antennas from the same tag to determine whether the RFID tag belongs to the lower, middle or upper layer (i.e. Level 1, Level 2, Level 3), thereby obtaining its relative position.

[0020] Only three levels of differentiation are considered here (i.e., Level 1, Level 2, and Level 3). For shelves in the warehouse, tags on the fourth layer and below are easily affected by record loss because they are far away from the main beam of the reader antenna, especially when the antenna is tilted. To locate multi-layer shelves (more than three layers), it is only necessary to scan at multiple heights (different positions). From the perspective of algorithm design, three-level differentiation is practical and efficient for shelf positioning in warehouses. The RFID positioning system installed on a drone or robot does not need to scan each shelf layer, which saves more time and reduces power consumption. The vertical positioning information of the tag is thus obtained.

[0021] The measured phase is used to calculate the horizontal positioning information of the tag, and the measured phase of the RFID tag is set to follow the Gaussian distribution Ψ m ~N(μ,σ 2 ), and each measurement phase is independent of each other. The horizontal positioning problem is solved by using variant-based maximum likelihood estimation. The horizontal position of the target label can be given as

[0022]

[0023]

[0024] Among them, (ψ m [n]) represents the measured phase of the nth tag, mod(·) is the modulo 2π operator, λ represents the signal wavelength, ψ 0represents the phase shift caused by the antenna and the tag, σ 2 is the statistical variance of the measured phase, d[n] = ||P a -P t || is from the antenna position P a =(X a , Y a , Z a ) to label position P t =(X t , Y t , Z t ) distance, P a [n]=(X a , Y a [n],Z a ), which indicates the position of the corresponding antenna when measuring the nth tag, where X a and Z a is a constant, Y a [n] is related to the motion state of the antenna;

[0025] Since only the horizontal position of the label is considered, the above formula can be written as

[0026]

[0027] Following the above technical solution, due to label diversity, label direction and frequency diversity, the phase shift ψ 0 will vary and cannot be calibrated in advance, but for a given tag and antenna, even if there are small fluctuations due to polarization mismatch as the antenna moves along the tag, ψ 0 can also be regarded as a constant, for which differential elimination is introduced to alleviate ψ 0 The influence of m [n] and the selected reference point ψ m [r] (select the first measured phase as reference) phase difference, that is

[0028] definition The above formula is converted to

[0029]

[0030] in The phase shift Ψ 0 be eliminated.

[0031] Following the above technical solution, the naive likelihood function matches well with the sine likelihood function defined by the sine function using the Taylor series approximation method, so the positioning estimate is

[0032]

[0033] From this we get the horizontal positioning information y tag .

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention uses an UHF radio frequency reader mounted on a drone to identify product packaging, thereby realizing rapid positioning of product packaging and laying a solid foundation for realizing intelligent warehousing. The present invention includes a radio frequency reader device mounted on a drone, packaging identification based on attached RFID tags, and a hybrid relative passive RFID positioning method based on received signal strength indicator (RSSI) and phase measurement. The present invention can attach RFID tags to goods, and then use RFID readers to identify and track the goods, thereby realizing full-process real-time monitoring and management of the goods. The present invention uses an RFID reader mounted on a drone to preliminarily realize intelligent warehousing management, such as rapid positioning of goods in the warehouse, inventory counting of goods, and safety of goods transportation. At the same time, the present invention can help enterprises improve the efficiency of goods transportation and reduce labor costs in the warehouse management process. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a structural block diagram of the intelligent identification and positioning system proposed by the present invention;

[0037] Figure 2 This is a schematic diagram of the RSSI and phase mixing algorithm proposed in the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0039] Embodiment 1:

[0040] The product packaging intelligent identification and positioning system based on drones proposed in this solution includes an intelligent identification and positioning software module and a remote monitoring module; among which:

[0041] The intelligent identification and positioning software module is used for the drone to collect UHF tag information during cruising and the host computer to analyze and judge the collected data, and finally realizes the positioning of product packaging through the relative positioning algorithm of RSSI and phase mixing; it includes a navigation real-time processing unit, a radio frequency tag information acquisition unit, and a product packaging intelligent identification and positioning unit;

[0042] The remote monitoring module is used to upload the data analyzed and processed by the intelligent identification and positioning software module from the host computer to the Alibaba Cloud database, and finally display the packaging number, packaging status parameters, packaging location, storage records and logistics information on the Internet of Things platform to achieve information sharing and remote control;

[0043] The navigation real-time processing unit checks the status of the radio frequency reader on board the drone before flight to ensure its normal operation, and formulates a flight plan and executes the flight mission according to the specific environment of the warehouse. During this process, feedback and adjustment of the flight status are carried out at any time;

[0044] The radio frequency tag information acquisition unit transmits information between the tag and the reader through a single-station backscatter link to collect tag information, unwrapping phase, received signal strength (RSSI), operating frequency and laser ranging information on the product packaging;

[0045] The product packaging intelligent identification and positioning unit includes packaging identity recognition, RSSI and phase mixing relative positioning algorithm.

[0046] The drone carries antenna No. 1 and antenna No. 2 that can rotate up and down. Antenna No. 1 rotates upward, and antenna No. 2 rotates downward. Antenna No. 1 and antenna No. 2 are used to collect RSSI information and phase information. Preliminary positioning information is obtained through the relative positioning algorithm mixed with RSSI and phase. At the same time, interference compensation is set to compensate and adjust the obtained preliminary information (vertical positioning information and horizontal positioning information) to obtain more accurate packaging positioning information;

[0047] Specifically, it uses the RSSI information obtained by the first antenna and the second antenna to obtain the vertical positioning information through calculation, and uses the phase information obtained by the first antenna and the second antenna to obtain the horizontal positioning information through calculation.

[0048] The information obtained by the antenna No. 1 includes: RSSI 1 The information obtained by antenna No. 1 and No. 2 includes: RSSI 2 and Phase II, compares the RSSI profile of the same tag from the antenna, RSSI caused by rotating the antenna 1 With RSSI 2 The difference is used to distinguish the vertical position of shelf items. However, due to the lateral polarization mismatch, it is necessary to compare the values ​​in the neighborhood of each tag RSSI peak value and define the neighborhood. where γ 0 is the RSSI peak index when the antenna is located directly in front of the tag (the RSSI peak index of each tag is different, but it is a constant for a specific tag), and ε represents the neighborhood Γ εHalf the length, therefore, if we compare RSSI profiles only within the defined RSSI neighborhood, the transverse polarization mismatch will have no effect. Furthermore, taking into account the measurement noise as well as the residual errors caused by channel fading and polarization mismatch, the judgment condition can be defined as:

[0049] Level 1:

[0050] Level 2:

[0051] Level 3:

[0052] When the above residual error is taken into account, Φ is the given RSSI offset threshold.

[0053] In the above scheme, the drone is equipped with a reader and works at any height in the warehouse. It scans three shelves within the range each time and compares the RSSI difference of the dual antennas from the same tag to determine whether the RFID tag belongs to the lower, middle or upper layer (i.e. Level 1, Level 2, Level 3), thereby obtaining its relative position.

[0054] Only three levels of differentiation are considered here (i.e., Level 1, Level 2, and Level 3). For shelves in the warehouse, tags on the fourth layer and below are easily affected by record loss because they are far away from the main beam of the reader antenna, especially when the antenna is tilted. To locate multi-layer shelves (more than three layers), it is only necessary to scan at multiple heights (different positions). From the perspective of algorithm design, three-level differentiation is practical and efficient for shelf positioning in warehouses. The RFID positioning system installed on a drone or robot does not need to scan each shelf layer, which saves more time and reduces power consumption. The vertical positioning information of the tag is thus obtained.

[0055] The measured phase is used to calculate the horizontal positioning information of the tag, and the measured phase of the RFID tag is set to follow the Gaussian distribution Ψ m ~N(μ,σ 2 ), and each measurement phase is independent of each other. The horizontal positioning problem is solved by using variant-based maximum likelihood estimation. The horizontal position of the target label can be given as

[0056]

[0057] Among them, (ψ m [n]) represents the measured phase of the nth tag, mod(·) is the modulo 2π operator, λ represents the signal wavelength, ψ 0 represents the phase shift caused by the antenna and the tag, σ 2 is the statistical variance of the measured phase, d[n] = ||Pa -P t || is from the antenna position P a =(X a , Y a , Z a ) to label position P t =(X t , Y t , Z t ) distance, P a [n]=(X a , Y a [n],Z a ), which indicates the position of the corresponding antenna when measuring the nth tag, where X a and Z a is a constant, Y a [n] is related to the motion state of the antenna;

[0058] Since only the horizontal position of the label is considered, the above formula can be written as

[0059]

[0060] Following the above technical solution, due to label diversity, label direction and frequency diversity, the phase shift ψ 0 will vary and cannot be calibrated in advance, but for a given tag and antenna, even if there are small fluctuations due to polarization mismatch as the antenna moves along the tag, ψ 0 can also be regarded as a constant, for which differential elimination is introduced to alleviate ψ 0 The influence of m [n] and the selected reference point ψ m [r] (select the first measured phase as reference) phase difference, that is

[0061] definition The above formula is converted to

[0062]

[0063] in The phase shift Ψ 0 be eliminated.

[0064] Following the above technical solution, the naive likelihood function matches well with the sine likelihood function defined by the sine function using the Taylor series approximation method, so the positioning estimate is

[0065]

[0066] Thus we get more accurate horizontal positioning information y tag

[0067] This solution uses an UHF radio frequency reader mounted on a drone to identify product packaging, thereby quickly locating product packaging and laying a solid foundation for realizing intelligent warehousing. This solution also includes a radio frequency reader device mounted on a drone, package identification based on attached RFID tags, and a hybrid relative passive RFID positioning method based on received signal strength indicator (RSSI) and phase measurement. This solution can attach RFID tags to goods, and then use RFID readers to identify and track the goods, thereby realizing full-process real-time monitoring and management of the goods. At the same time, the use of drones equipped with RFID readers can initially realize intelligent warehousing management, such as rapid positioning of goods in warehouses, inventory counting of goods, and safety of goods transportation. At the same time, the present invention can help enterprises improve the efficiency of goods transportation and reduce labor costs in the warehouse management process.

[0068] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. The intelligent identification and positioning system for product packaging based on drones is characterized by: It includes intelligent identification and positioning software module and remote monitoring module; among which: The intelligent identification and positioning software module is used for the drone to collect UHF tag information during cruising and the host computer to analyze and judge the collected data, and finally realizes the positioning of product packaging through the relative positioning algorithm of RSSI and phase mixing; it includes a navigation real-time processing unit, a radio frequency tag information acquisition unit, and a product packaging intelligent identification and positioning unit; The remote monitoring module is used to upload the data analyzed and processed by the intelligent identification and positioning software module from the host computer to the Alibaba Cloud database, and finally display the packaging number, packaging status parameters, packaging location, storage records and logistics information on the Internet of Things platform to achieve information sharing and remote control; The navigation real-time processing unit checks the status of the radio frequency reader on board the drone before flight to ensure its normal operation, and formulates a flight plan and executes the flight mission according to the specific environment of the warehouse. During this process, feedback and adjustment of the flight status are carried out at any time; The radio frequency tag information acquisition unit transmits information between the tag and the reader through a single-station backscatter link to collect tag information, unwrapping phase, RSSI, operating frequency and laser ranging information on the product packaging; The product packaging intelligent identification and positioning unit includes packaging identity recognition, RSSI and phase hybrid relative positioning algorithm; The drone carries antenna No. 1 and antenna No. 2 that can rotate up and down. Antenna No. 1 and antenna No. 2 are used to collect RSSI information and phase information. The preliminary positioning information is obtained through the relative positioning algorithm mixed with RSSI and phase. At the same time, interference compensation is set to compensate and adjust the preliminary information to obtain more accurate packaging positioning information; Specifically, it uses the RSSI information obtained by the first antenna and the second antenna to obtain vertical positioning information through calculation, and uses the phase information obtained by the first antenna and the second antenna to obtain horizontal positioning information through calculation; The information obtained by antenna No. 1 includes: RSSI1 and phase 1, and the information obtained by antenna No. 2 includes: RSSI2 and phase 2. The RSSI profiles of the same tag from the antennas are compared. The difference between RSSI1 and RSSI2 caused by rotating the antenna is used to distinguish the vertical positions of the items on the shelf. However, due to the lateral polarization mismatch, it is necessary to compare the values ​​in the neighborhood of the RSSI peak of each tag and define the neighborhood. Where γ0 is the RSSI peak index when the antenna is located directly in front of the tag, ε represents the neighborhood Γ ε Half the length, therefore, if we compare RSSI profiles only within the defined RSSI neighborhood, the transverse polarization mismatch will have no effect. Furthermore, taking into account the measurement noise as well as the residual errors caused by channel fading and polarization mismatch, the judgment condition can be defined as: When the above residual error is considered, Φ is the given RSSI offset threshold; The drone is equipped with a reader and works at any height in the warehouse. Each scan covers three shelves. By comparing the RSSI difference between the dual antennas of the same tag, it can be determined whether the RFID tag belongs to the lower, middle or upper layer, i.e. Level 1, Level 2, Level 3, and thus its relative position.

2. The product packaging intelligent identification and positioning system based on drones as claimed in claim 1, characterized in that: The measured phase is used to calculate the horizontal positioning information of the tag, and the measured phase of the RFID tag is set to follow the Gaussian distribution Ψ m ~N(μ,σ 2 ), and each measurement phase is independent of each other. The horizontal positioning problem is solved by using variant-based maximum likelihood estimation. The horizontal position of the target label can be given as Among them, (ψ m [n]) represents the measured phase of the nth tag, mod(·) is the modulo 2π operator, λ represents the signal wavelength, ψ0 represents the phase offset caused by the antenna and the tag, and σ 2 is the statistical variance of the measured phase, d[n] = ||P a [n]-P t || is from the antenna position P a [n]=(X a , Y a [n],Z a ) to label position P t =(X t , Y t , Z t ) distance, P a [n]=(X a , Y a [n],Z a ), which indicates the position of the corresponding antenna when measuring the nth tag, where X a and Z a is a constant, Y a [n] is related to the motion state of the antenna; Since only the horizontal position of the label is considered, the above formula can be written as Following the above technical solution, due to tag diversity, tag direction and frequency diversity, the phase shift ψ0 will change and cannot be calibrated in advance. Therefore, differential elimination is introduced to mitigate the impact of ψ0, which uses each sampling position ψ m [n] and the selected reference point ψ m The phase difference between [r], that is, definition The above formula is converted to in Thereby the phase offset Ψ0 is eliminated; Following the above technical solution, the naive likelihood function matches well with the sine likelihood function defined by the sine function using the Taylor series approximation method, so the positioning estimate is From this we get the horizontal positioning information y tag .

Images