Positioning method and system of smart shopping cart, computer equipment, and storage medium

Abstract

To reduce complexity of hardware implementation and a maintenance cost of positioning technology, and improve positioning accuracy and stability.SOLUTION: A positioning method of a smart shopping cart includes the steps of: constructing the positioning fingerprint database of the smart shopping cart on the basis of commodity attributes collected by a smart device and price tag heartbeat signals received by a communication module; when movement data of a main body is collected by a motion sensor, sending, by the communication module, all price tag heartbeat signals received in a second predetermined time window to a server; and matching, by the server, all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to acquire the positioning information of the smart shopping cart.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to the field of positioning technology, and in particular to a positioning method, system, computer device and storage medium for a smart shopping cart. [Background technology]

[0002] At present, all large shopping malls around the world are undergoing digital upgrades. In the application of digitalization, electronic shelf labels have replaced traditional paper price tags. In addition to displaying ordinary information, electronic shelf labels are also used for high-speed picking, stockout management, high-speed inventory, and human-computer interaction with users. When a shopping cart travels through a shelf area, it is usually necessary to locate the shopping cart, navigate the shopping cart, and provide recommended services based on the shopping cart's location. Currently, Bluetooth (registered trademark) positioning technology is typically used to determine the position of shopping carts, but in order to cover all shelf aisles in large shopping malls, a huge number of Bluetooth beacons must be deployed, which poses problems such as complex implementation and high maintenance costs. Summary of the Invention [Problem to be solved by the invention]

[0003] In order to solve the problems in the prior art, such as the complex implementation and high maintenance costs of the shopping cart positioning methods in the prior art, the present invention provides a method, system, computer equipment and storage medium for positioning a smart shopping cart, which realizes a fingerprint positioning method for a shopping cart using an electronic price tag system installed in a store, reduces the complexity of the hardware implementation of the positioning technology and the maintenance costs, and improves the positioning accuracy and stability. [Means for solving the problem]

[0004] The present invention according to a first aspect is a positioning method for a smart shopping cart, which is applied to a smart shopping cart in a store in which an electronic price tag system including an electronic price tag with known position information that transmits heartbeat packets and a server is installed, the smart shopping cart including a main body, a communication module provided in the main body for receiving the heartbeat packets, a motion sensor for collecting movement data, and a smart device for collecting product attributes, the positioning method for the smart shopping cart includes the steps of: when product attributes of an arbitrary product are collected by the smart device, transmitting all heartbeat packets including a price tag heartbeat signal received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the electronic price tag system; and transmitting the product attributes to the server. obtain target location information of the product based on the product attributes; associate the target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window and an RSSI value at which each heartbeat packet is received to obtain fingerprint data corresponding to the target location information; and construct a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained; the communication module transmits all price tag heartbeat signals received in a second predetermined time window to a server when the movement data of the main body is collected by the motion sensor; and the server matches all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0005] Preferably, the price tag heartbeat signal includes a price tag ID and a reporting time, and the fingerprint data corresponding to the target location information includes a plurality of price tag IDs and one RSSI weighted average value corresponding to each price tag ID.

[0006] Preferably, the first predetermined time window is the sum of a time length T1 until product attributes are collected by the smart device, a time length T2 from when the product attributes are collected by the smart device until the shopping cart starts moving, and a time length T3 from when the shopping cart starts moving, and the second predetermined time window is a time length T4 until the movement data is collected by the motion sensor.

[0007] Preferably, the step of the server acquiring target location information of the product based on the product attributes includes a step of, when the server acquires a plurality of location information based on the product attributes, sorting the plurality of location information based on positioning information at a time immediately prior to the shopping cart, and acquiring the target location information of the product.

[0008] Preferably, after constructing a positioning fingerprint database corresponding to the store, when the smart device continues to collect product attributes, the method further includes the steps of obtaining fingerprint data corresponding to the location where the product is located, and updating the positioning fingerprint database using the fingerprint data corresponding to the location where the product is located.

[0009] Preferably, before constructing a positioning fingerprint database corresponding to the store, the server further includes the steps of: acquiring a shelf number and a shelf section index corresponding to each electronic price tag based on the price tag ID in all price tag heartbeat signals received in the second predetermined time window; aggregating and aggregating price tag heartbeat signals of the same shelf section based on the shelf number and the shelf section index to obtain an aggregated indicator corresponding to each shelf section; and comprehensively analyzing the aggregated indicators corresponding to each shelf section to determine a target shelf section and set the coordinate position of the target shelf section as positioning information of the smart shopping cart.

[0010] Preferably, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, before constructing a positioning fingerprint database corresponding to the store, the method further includes a step of calculating an azimuth angle of the emission signal source of each electronic price tag based on baseband signal characteristics of the predetermined sequence signal received by each antenna, and a step of calculating positioning information of the smart shopping cart based on position information of at least three non-collinear electronic price tags and the azimuth angles corresponding to the at least three non-collinear electronic price tags.

[0011] Preferably, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, the server matches all price tag heartbeat signals received in a second predetermined time window with the positioning fingerprint database to obtain the positioning information of the smart shopping cart, and then further includes a step of calculating the azimuth angle of the emitted signal source of each electronic price tag based on the baseband signal characteristics at which the predetermined sequence signal is received by each antenna, and a step of correcting the positioning information of the smart shopping cart based on the azimuth angle to obtain the target positioning information of the smart shopping cart.

[0012] Preferably, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, when product attributes of any product are collected by the smart device, the method further includes the steps of: sending all price tag heartbeat signals received by the communication module in a first predetermined time window, the azimuth angle at which each price tag heartbeat signal is received, and the said product attributes to a server in the electronic price tag system; the server obtains position information corresponding to the product based on the product attributes, and obtains position information corresponding to the price tag based on the price tag ID in the price tag heartbeat signal; storing the position information and the azimuth angle corresponding to the price tag in an input dataset, and storing the position information corresponding to the product in an output dataset; training a machine learning algorithm based on the input dataset and the output dataset to obtain a positioning model of the shopping cart; and inputting the price tag position information and the azimuth angle corresponding to the currently received price tag heartbeat signal into the positioning model of the shopping cart for identification by the communication module, and obtaining current positioning information of the smart shopping cart.

[0013] The present invention according to a second aspect is a positioning system for a smart shopping cart, comprising an electronic price tag having known position information, a server, and a smart shopping cart, the smart shopping cart comprising a main body, a communication module provided in the main body for receiving heartbeat packets, a motion sensor for collecting movement data, and a smart device for collecting product attributes, the electronic price tag for transmitting heartbeat packets including a price tag heartbeat signal, the communication module for transmitting all heartbeat packets received in a first predetermined time window and the product attributes to a server in the electronic price tag system when product attributes of any product are collected by the smart device, and the server determines a target position of the product based on the product attributes. and acquiring target location information of the product, associating target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window to obtain fingerprint data corresponding to the target location information, and constructing a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained, the communication module is further configured to transmit all price tag heartbeat signals received in a second predetermined time window to a server when movement data of the main body is collected by the motion sensor, and the server is further configured to match all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0014] The present invention according to a third aspect is a computer device including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein when product attributes of an arbitrary product are collected by the smart device by executing the computer program by the processor, the computer device includes a step of transmitting all heartbeat packets including a price tag heartbeat signal received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the electronic price tag system; the server obtains target location information of the product based on the product attributes, and transmits the target location information of the product and the RSSI value of the communication module in the first predetermined time window to a server in the electronic price tag system; the communication module associates all price tag heartbeat signals received by the communication module with the RSSI value at which each heartbeat packet is received to obtain fingerprint data corresponding to the target location information, and constructs a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained; the communication module transmits all price tag heartbeat signals received in a second predetermined time window to a server when movement data of the main body is collected by the motion sensor; and the server matches all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0015] The present invention according to a fourth aspect is a readable storage medium having a computer program stored thereon, the computer program being executed by the processor to transmit, when product attributes of any product are collected by the smart device, all heartbeat packets including price tag heartbeat signals received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the electronic price tag system; the server obtains target location information of the product based on the product attributes, and transmits the target location information of the product, all price tag heartbeat signals received by the communication module in the first predetermined time window, and each A readable storage medium having a computer program stored thereon, the computer program being characterized by realizing the steps of: associating the RSSI value at which the heartbeat packet is received to obtain fingerprint data corresponding to the target location information, and constructing a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained; the communication module transmitting all price tag heartbeat signals received in a second predetermined time window to a server when the movement data of the main body is collected by the motion sensor; and the server matching all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart. Effect of the Invention

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention uses an electronic price tag system installed in a store to build a positioning fingerprint database in which fingerprint data is associated with the position of each product based on product attributes collected by a smart device installed in the shopping cart and the price tag heartbeat signal received by the communication module, and when locating the shopping cart, matches the collected price tag heartbeat signal with the positioning fingerprint database to realize real-time positioning of the shopping cart.Therefore, the electronic price tag system installed in a store realizes a fingerprint positioning method for the shopping cart, reduces the complexity and maintenance cost of hardware implementation of the positioning technology, improves positioning accuracy and stability, and has a wide range of applications. [Brief description of the drawings]

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

[0018] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments are described below clearly and completely with reference to the drawings in the embodiments of the present application, and it is clear that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. According to the embodiments of the present application, all other embodiments obtained by a person skilled in the art without inventive ideas are within the scope of the present application. In a first aspect, the present invention provides a positioning method for a smart shopping cart, specifically including the following embodiments:

[0019] Example 1 Fig. 1 is a flow chart showing a first type of smart shopping cart positioning method according to an embodiment of the present invention. As shown in Fig. 1, the smart shopping cart positioning method specifically includes: when the product attributes of any product are collected by the smart device, sending all heartbeat packets including price tag heartbeat signals received by the communication module in a first predetermined time window, the RSSI value at which each heartbeat packet is received, and the product attributes to a server in an electronic price tag system in step S101.

[0020] The positioning method for a smart shopping cart according to this embodiment is applied to a smart shopping cart in a store in which an electronic price tag system is installed, the electronic price tag system including an electronic price tag with known position information that transmits heartbeat packets, and a server, the electronic price tag system further including a base station, and the electronic price tag communicates with the server via the base station, as shown in Fig. 2. When constructing the electronic price tag system, positioning is performed for each electronic price tag in the store with accuracy up to the shelf level based on the neighborhood learning positioning algorithm in the prior art, that is, the position information of each electronic price tag in this embodiment is known. The server in Figure 2 stores the location information of each EPL tag and locates the EPL tag and smart shopping cart using the neighborhood data and heartbeat data; the server manages the behavior of the EPL tag via the base station; the base station directly connects to the EPL tag and is responsible for transmitting messages and data; the EPL tag periodically broadcasts a heartbeat and, upon receiving a positioning command, performs a patrol inspection and collects the signals of neighboring price tags as neighborhood information; the shopping cart receives and caches the heartbeats of the EPL tags, and when positioning is required, it can use the cached data to locate.

[0021] The mounting forms of EPLs on shelves are as follows, as shown in Fig. 3: 1. Placed horizontally on a flat surface, 2. Hanged on a hook, and 3. Hanged diagonally on a tiered board. Each EPL transmits heartbeat packets in the broadcast frequency band at regular time intervals, and the 2.4G receiving module can receive the heartbeat packets in the broadcast frequency band.

[0022] As shown in Fig. 4, the smart shopping cart in this embodiment includes a main body 4, a communication module 5 installed in the main body for receiving heartbeat packets, a motion sensor 6 for collecting movement data, and a smart device 7 for collecting product attributes, and the installation locations of the communication module 5, the motion sensor 6, and the smart device 7 are not limited to those shown in Fig. 4. The communication module 5 is a 2.4G communication module, and when a user pushes the shopping cart to purchase products in a shopping mall, the communication module can receive heartbeat packets sent from price tags on surrounding shelves, and the smart device 7 can scan product barcodes and automatically enter them into a purchase order, and the smart device can connect to a server in the background. When the shopping cart is moved, the motion sensor 6 can detect the movement of the shopping cart and transmit the motion information to the smart device.

[0023] In this embodiment, before performing fingerprint positioning on the shopping cart, a fingerprint database associated with the store must be constructed. When a customer purchases a product or a store staff takes the trouble to collect product attributes, the barcode of the product is scanned by a smart device attached to the shopping cart, while triggering a communication module to collect heartbeat packets transmitted from nearby electronic price tags within a certain time, and all of the heartbeat data and the product attributes are reported to the server. In order to collect as many heartbeat packets as possible for identifying the location, the certain time is the above-mentioned first predetermined time window, and the definition of the first predetermined time window is as shown in FIG. 5. The first predetermined time window is the sum of the time length T1 until the product attributes are collected by the smart device, the time length T2 from the collection of the product attributes by the smart device to the start of the shopping cart moving, and the time length T3 from the start of the shopping cart moving, where the time length T2 is the time length that the shopping cart stays at the original location after the product is scanned by the smart device, and if the product is scanned, the time length T3 is the time length that the shopping cart stays at the original location after the product is scanned by the smart device. If the shopping cart does not stay there after that, the time length T2 is 0. In this embodiment, the time length T1 and the time length T3 each have a value of 10S, that is, the collected heartbeat packets are divided into three parts: a heartbeat packet received 10s before triggering the product scan, a heartbeat packet received in the period Ns from triggering the product scan to the motion sensor detecting that the shopping cart has moved again, and a heartbeat packet received 10 seconds after the shopping cart has moved again, and the heartbeat packet includes the price tag heartbeat signal.

[0024] In this embodiment, the price tag heartbeat signal includes a price tag ID and a reporting time, and the price tag sets the heartbeat transmission power to 0dbm to ensure normal brushing. In the positioning of the shopping cart, due to this installation mode, far-away price tags are also collected by the shopping cart, which affects the construction of the fingerprint library. Therefore, before reporting the data, the smart device screens the RSSI value of the price tag heartbeat signal by setting a threshold value. That is, to ensure that it is a nearby price tag, the heartbeat of this price tag is reported to the server only if the signal strength exceeds a certain threshold value.

[0025] Step S102: The server obtains target location information of the product based on the product attributes, associates the target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window and the RSSI value at which each heartbeat packet is received as fingerprint data corresponding to the target location information, and builds a positioning fingerprint database corresponding to the store after obtaining the fingerprint data of a predetermined proportion of the store.

[0026] In this embodiment, the server determines the target location information where the product is located based on the received product attributes, and the target location information is determined by the association relationship between the product and the price tag, that is, the location information of the associated electronic price tag is used as the target location information, and the server can obtain fingerprint data corresponding to the target location information by associating the target location information with all price tag heartbeat signals received in a first predetermined time window, and the heartbeat signal corresponding to the target location information includes multiple price tag heartbeat signals transmitted multiple times from the same electronic price tag, and the RSSI values ​​of the price tag heartbeat signals received multiple times may be the same or different, so it is necessary to take a weighted average of the multiple RSSI values ​​of the multiple price tag heartbeat signals transmitted from the same electronic price tag based on the price tag ID, so the fingerprint data corresponding to the target location information includes the price tag ID and the weighted average value of the RSSI. When a predetermined percentage of the fingerprint data is obtained, it is indicated that the establishment of the positioning fingerprint database is completed, and the predetermined percentage may be any percentage such as 70%, 80% of the entire fingerprint data. It should be noted that for the fingerprint positioning of the shopping cart, the entire store is divided into several grids, and in this scene, after the product is scanned by the user, the position of the product will correspond to the grid.

[0027] The construction of the positioning fingerprint database has several features: (1) A time threshold needs to be set, and only heartbeat data within this time threshold is used to construct the fingerprint. This is to prevent the price tag from being moved and affecting the reliability of the fingerprint library. The heartbeat data within the time threshold needs to be divided into several parts based on the time they are obtained, and the closer the heartbeat data is to the current time, the higher the reliability is, and the corresponding weight is assigned. The weight assignment for each heartbeat data can be done based on the following formula:

[0028]

number

[0029] where x is the number of allocated times and the maximum weight value, t is the time threshold, and t c is the current time, t r is the reporting time of the heartbeat packet. The RSSI weighted average of identical price tags corresponding to a single location in the fingerprint library can be calculated based on the following formula:

[0030]

number

[0031] Here, RSSI w_avg is the weighted average of the RSSI of the price tag corresponding to the position in the fingerprint library, and w i is the weight value corresponding to the i-th heartbeat data, and RSSI i is the RSSI value corresponding to the i-th heartbeat data. When performing fingerprint matching, if heartbeat data is received multiple times from the same price tag at the same location, a weighted average value may be calculated for the heartbeat data according to a formula. When calculating the similarity of fingerprints, the Euclidean distance can be used.

[0032]

number

[0033] Here, n represents the weighted average of n price tags in the area, and RSSI RX_i represents the RSSI value of the currently received price tag, and the larger the distance, the smaller the similarity. (2) The fingerprint data of a location includes three parts: price tag ID, RSSI value, and reporting time, and is finally sorted based on the time threshold. The fingerprint data to build the fingerprint database includes price tag ID and RSSI weighted average value. (3) As the user continuously collects products, the fingerprint database is continuously updated. (4) The fingerprint library includes two dimensions: fingerprints at a single position and full-field fingerprints. 1) For fingerprints at a single position, multiple products may be associated with that position. Only when the proportion or the number of products collected at that position exceeds a certain threshold can that position be considered reliable. 2) For full-field fingerprints, when the proportion of reliable positions exceeds a certain threshold, it is considered that the construction of the full-field fingerprint is basically completed. (5) If there is no fingerprint corresponding to a position, the shopping cart cannot be positioned at that position. Regarding the weight of the heartbeat data, according to the actual scene, the weight of the heartbeat data within the time threshold t can be made equal to 1, and the weight of the heartbeat data outside the time threshold t can be made equal to 0. At this time, the weighted average value of RSSI is the average value of RSSI.

[0034]

Number

[0035] The Weight formula can be adjusted according to the actual situation. When calculating the similarity of fingerprints, according to the actual scene, in addition to the method using Euclidean distance, other calculation methods such as the following may also be used.

[0036]

Number

[0037] In another embodiment of the present invention, the positioning method for the mart shopping cart further includes, after constructing a positioning fingerprint database corresponding to the store, when the smart device continues to collect product attributes, a step of obtaining fingerprint data corresponding to the location where the product is located, and a step of updating the positioning fingerprint database using the fingerprint data corresponding to the location where the product is located.

[0038] Note that updating the positioning fingerprint database includes adding new fingerprint data and updating constructed fingerprint data.

[0039] In yet another embodiment of the present invention, the step of the server acquiring target location information of the product based on the product attributes includes a step of, when the server acquires a plurality of location information based on the product attributes, sorting the plurality of location information based on positioning information at a time immediately prior to the shopping cart, and acquiring the target location information of the product.

[0040] In addition, during the actual operation of a store, one product may be placed in multiple locations. For example, laundry liquid may be placed not only on a normal item storage shelf but also in a promotional area. As a result, the scanned product attributes correspond to multiple locations in the store. Therefore, when the server obtains multiple location information based on the product attributes, it selects the multiple location information based on the positioning information of the shopping cart at the previous time to obtain the target location information of the product. In other words, among the multiple location information, the location information closest to the positioning information at the previous time is set as the target location information.

[0041] Step S103: The communication module sends all price tag heartbeat signals received in a second predetermined time window to a server when movement data of the body of the shopping cart is collected by the motion sensor.

[0042] Step S104: The server matches all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain the positioning information of the smart shopping cart.

[0043] Here, the second predetermined time window is a time length T4 until the movement data is collected by the motion sensor.

[0044] In addition, after establishing a positioning fingerprint database corresponding to the store, when the shopping cart is pushed, the real-time positioning function of the shopping cart can be triggered. When the motion sensor collects the movement data of the body of the shopping cart, the smart device in the shopping cart can send all price tag heartbeat signals received in the most recent time window to a server, and the server positions the shopping cart in real time based on a big data matching method or a machine learning algorithm.

[0045] In one embodiment of this embodiment, the big data matching method is a traditional Bayesian fingerprint location method, and according to the received signal, a signal vector m is obtained, where m=(<esl_id1,rssi_avg1> ,< esl_id2,rssi_avg2>,…,< esl_idn,rssi_avgn>). Based on the positioning fingerprint database, all positions p are obtained, so the aim is to find the maximum probability P(p|m), i.e., if the signal vector is m, the probability of being located at position p is the maximum, and by performing a Bayes transformation on the equation, we obtain P(p|m)=P(m|P)*P(p) / P(m). At one sampling time, the probability that signal vector m is obtained by sampling is a constant, so P(p|m)=P(m|P)*P(p). In this scene, by analyzing the data, P(p|m) and P(p) can be obtained, and the position corresponding to the most likely signal vector can be found.

[0046] In another embodiment of this embodiment, the machine learning algorithm is KNN algorithm (K Nearest Neighbors), which is a simple machine learning algorithm, the principle of which is to calculate the coordinates of the collected location based on the k locations closest to the collected signal vector m. Here, the two key points are: 1) to select the optimal k value according to the positioning fingerprint database; 2) to calculate the distance between m and the vector fingerprint of other locations by methods such as Euclidean distance, Manhattan distance or Chebyshev distance.

[0047] Compared with the prior art, this embodiment has the following beneficial effects. The present invention uses an electronic price tag system installed in a store to build a positioning fingerprint database in which fingerprint data is associated with each product position based on product attributes collected by a smart device in the shopping cart and the price tag heartbeat signal received by the communication module, and when locating the shopping cart, matches the collected price tag heartbeat signal with the positioning fingerprint database to realize real-time positioning of the shopping cart.Therefore, the electronic price tag system installed in a store realizes a fingerprint positioning method for the shopping cart, reduces the complexity and maintenance cost of hardware implementation of the positioning technology, improves positioning accuracy and stability, and has a wide range of applications.

[0048] Example 2 6 is a flow chart showing a second type of smart shopping cart positioning method according to an embodiment of the present invention. As shown in FIG. 6, the positioning of the shopping cart is performed before building a positioning fingerprint database corresponding to the store. The server obtains a shelf number and a shelf section index corresponding to each EPL based on the price tag ID in all price tag heartbeat signals received in the second predetermined time window in step S201; a step S202 of aggregating and tallying price tag heartbeat signals related to the same shelf section based on the shelf number and the shelf section index to obtain a tally index corresponding to each shelf section; The method further includes a step S203 of determining a target shelf section by comprehensively analyzing the tabulated indicators corresponding to each shelf section, and setting the coordinate position of the target shelf section as positioning information of the smart shopping cart.

[0049] In addition, when the shopping cart departs to perform positioning and the positioning fingerprint database has not been established, the smart device of the shopping cart will send the heartbeat data received in the most recent time window to the server, and the server can position the shopping cart through a special positioning process.

[0050] In this embodiment, a special positioning process is to store the positioning position, shelf number and shelf section index of each EPL in the database of the server. The smart device of the shopping cart can calculate and identify the aisle where the shopping cart is currently located based on the heartbeat data collected within a certain time window. The specific algorithm process is as follows: (1) perform pre-processing on the heartbeat data uploaded from the shopping cart, including removing outliers, noise and duplicate data, to improve data quality and accuracy; (2) perform EPL information query processing, which queries the shelf number and shelf section index corresponding to the EPL in the server based on the identification ID of the EPL in the heartbeat data of the shopping cart; (3) perform aggregation and tabulation on the heartbeat data of the EPL on the shelf of the same shelf based on the queried shelf number and shelf section index, including calculating the average value, reciprocal average value, variance and other aggregation indicators of the heartbeat data of the shelf surface of each shelf, and determining the location information of the shopping cart. (4) Regarding aisle determination, the aggregated data is weighted and added, and a comprehensive analysis is performed according to the shelf position relationship to estimate the aisle where the shopping cart is located, and the shelf position relationship is determined based on the supermarket map and layout such as the relative positions and arrangement of shelves, where the comprehensive analysis includes sorting the aggregated indicators of each shelf section and determining the sorted shelf section at the top as the target shelf section. (5) A positioning result is returned, which returns information on the aisle where the shopping cart is currently located, thereby providing a real-time positioning service for smart shopping carts.

[0051] Example 3 7 is a flow chart showing a third type of smart shopping cart positioning method according to an embodiment of the present invention. As shown in FIG. 7, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, the positioning of the shopping cart includes the following steps before constructing a positioning fingerprint database corresponding to the store: S301, calculating an azimuth angle of a signal source emitted from each EPL based on baseband signal characteristics of the predetermined sequence signal received by each antenna; The method further includes a step S302 of calculating positioning information of the smart shopping cart based on position information of at least three non-collinear EPLs and azimuth angles corresponding to the at least three non-collinear EPLs.

[0052] In addition, each EPL attaches a signal of a predetermined known sequence to the end of the price tag heartbeat signal to be transmitted, so that the communication module of the shopping cart can easily calculate the azimuth angle when the communication module of the shopping cart receives the known sequence signal. The communication module with AOA positioning function in the shopping cart includes an antenna array consisting of multiple antennas, and the communication module can control the receiving time window of each antenna with high precision and calculate the azimuth angle of the transmitting signal source based on the baseband signal characteristics of the different antenna receiving signals. No matter which direction the head of the shopping cart faces, the position of the shopping cart can be calculated based on the known positions of the three non-collinear price tags, where the three non-collinear EPL tags indicate that the three EPL tags are not located on a straight line. As shown in FIG. 8, (X, Y) represents the positioning coordinates of the shopping cart, (X 1 ,Y 1 ) represents the position coordinates of the first electronic price tag, and (X 2 ,Y 2 ) represents the position coordinates of the second electronic price tag, and (X 3 ,Y 3 ) represents the position coordinates of the third electronic price tag, and α 1 represents the azimuth angle of the emission signal source of the first EPL, and α 2 represents the azimuth angle of the emission signal source of the second electronic price tag, and α 3 represents the azimuth angle of the emission signal source of the third electronic price tag, and R 1 → represents the direction vector from the first EPL to the shopping cart, and R 2 → represents the direction vector from the second EPL tag to the shopping cart, and R 3→ represents the direction vector from the third EPL tag to the shopping cart, and D 1 → represents a direction vector from the first EPL to the second EPL, and D 2 → represents a direction vector from the second EPL to the third EPL, and D 3 → represents a directional vector from the third EPL to the first EPL and has the following formula:

[0053]

number

[0054] α 1 , α 2 , α 3 , ||D 1 → ||, ||D 2 → ||, ||D 3 → Since || is known, ||R 1 → ||, ||R 2 → ||, ||R 3 → The value of || can be calculated by the following equation:

[0055]

number

[0056] The shopping cart coordinates can be calculated. In addition, for a shopping cart whose position has been determined by AOA, the fingerprint library may be updated using heartbeat data collected before and after the positioning (first predetermined time window), where the multiple antennas in the communication module include, but are not limited to, an antenna array in the form of a 6-antenna configuration in FIG. 9, an antenna array in the form of an 8-antenna configuration in FIG. 10, and an antenna array in the form of a 9-antenna configuration in FIG. 11.

[0057] Example 4 In this embodiment, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, the positioning method of the mart shopping cart includes: after the server matches all price tag heartbeat signals received in a second predetermined time window with the positioning fingerprint database to obtain the positioning information of the smart shopping cart; calculating an azimuth angle of a emitted signal source of each EPL based on baseband signal characteristics of the predetermined sequence signals received by each antenna; and correcting the positioning information of the smart shopping cart based on the azimuth angle to obtain target positioning information of the smart shopping cart.

[0058] 12, the semicircular arc indicated by the reference numeral 8 represents the direction in which the EPL transmits the heartbeat packet, the gray block indicated by the reference numeral 9 represents the EPL, the white rectangle indicated by the reference numeral 10 represents the shelf, the reference numeral 11 represents the shopping cart, and the reference numeral 12 represents the radiation range for receiving the signal from the communication module in the shopping cart, the direction in which the EPL transmits the signal is mainly toward the inside of the aisle and is not easily transmitted through the central partition plate of the shelf, the communication module in the shopping cart can receive the signal transmitted by the EPL, and the signal strength is within a certain range. Therefore, when the shopping cart is in the center of the aisle, it can be located in real time by the received heartbeat signal of the EPL.

[0059] In this embodiment, the communication module with AOA positioning function includes an antenna array consisting of multiple antennas. The communication module can precisely control the receiving time window of each antenna, and calculate the azimuth angle of the transmitting signal source based on the baseband signal characteristics of the reception of signals by different antennas. Based on the azimuth angle, the direction of the price tag relative to the shopping cart can be obtained. As shown in FIG. 13, the shopping cart's orientation, motion direction, etc. can be determined to obtain more accurate location information of the shopping cart. In FIG. 13, reference numeral 13 represents the aisle where the shopping cart is located, and reference numeral 14 represents the range where the shopping cart is located, which is determined based on the azimuth angle. When locating the shopping cart, due to the fluctuation of the signal strength, the positioning accuracy can only be determined to be located in the aisle, as shown in the gray part indicated by reference numeral 13, and after taking the azimuth angle into consideration, the positioning accuracy can be determined to be located in the light gray triangular area indicated by reference numeral 14, based on the positional relationship of the electronic price tag to the shopping cart, thereby improving the positioning accuracy.

[0060] Example 5 14 is a flow chart of a fourth type of smart shopping cart positioning method according to an embodiment of the present invention. As shown in FIG. 14, when the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, the mart shopping cart positioning method includes: S401, when the product attributes of any product are collected by the smart device, sending all price tag heartbeat signals received by the communication module in a first predetermined time window, the azimuth angle at which each price tag heartbeat signal is received, and the product attributes to a server in the EPL system; The server acquires location information corresponding to the product based on the product attributes, and acquires location information corresponding to the price tag based on the price tag ID in the price tag heartbeat signal in step S402; S403: storing the location information corresponding to the price tag and the azimuth angle in an input data set, and storing the location information corresponding to the item in an output data set; training a machine learning algorithm based on the input data set and the output data set to obtain a shopping cart positioning model; and The method further includes step S405, in which the communication module inputs the price tag position information and azimuth angle corresponding to the price tag heartbeat signal currently received into the positioning model of the shopping cart for identification, and obtains the current positioning information of the smart shopping cart.

[0061] In addition, the heartbeat data and recorded azimuth angle information sent by the smart shopping cart to the server are used to automatically train the fingerprint matching algorithm in addition to constructing the positioning fingerprint database, and the fingerprint matching algorithm can be optimized and trained by using the currently received heartbeat data and recorded azimuth angle information by the shopping cart as input, the matching result as output, and the location of the product as a label. When positioning the shopping cart in real time, the price tag position information and azimuth angle are obtained according to the currently received price tag heartbeat signal, and the price tag position information and azimuth angle are input into the trained fingerprint matching algorithm for identification, thereby obtaining the current positioning information of the shopping cart.

[0062] The present invention according to a second aspect is a positioning system for a smart shopping cart, comprising an electronic price tag whose position information is known, a server, and a smart shopping cart, the smart shopping cart comprising a main body, a communication module provided in the main body for receiving heartbeat packets, a motion sensor for collecting movement data, and a smart device for collecting product attributes, the electronic price tag for transmitting heartbeat packets including a price tag heartbeat signal, the communication module for transmitting all heartbeat packets received in a first predetermined time window and the product attributes to a server in the electronic price tag system when product attributes of any product are collected by the smart device, and the server for determining a target position of the product based on the product attributes. and acquiring information, associating target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window to obtain fingerprint data corresponding to the target location information, and constructing a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is acquired, the communication module is further configured to transmit all price tag heartbeat signals received in a second predetermined time window to a server when the movement data of the main body is collected by the motion sensor, and the server is further configured to match all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0063] The present invention according to a third aspect is a computer device including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein when product attributes of an arbitrary product are collected by the smart device by executing the computer program by the processor, the computer device includes a step of transmitting all heartbeat packets including a price tag heartbeat signal received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the electronic price tag system; the server obtains target location information of the product based on the product attributes, and transmits the target location information of the product and the RSSI value of the communication module in the first predetermined time window to a server in the electronic price tag system; the communication module associates all price tag heartbeat signals received by the communication module with the RSSI value at which each heartbeat packet is received to obtain fingerprint data corresponding to the target location information, and when a predetermined percentage of fingerprint data of a store is obtained, constructs a positioning fingerprint database corresponding to the store; when movement data of the main body is collected by the motion sensor, the communication module transmits all price tag heartbeat signals received in a second predetermined time window to a server; and the server matches all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0064] The present invention according to a fourth aspect is a readable storage medium having a computer program stored thereon, the computer program being executed by the processor to transmit, when product attributes of any product are collected by the smart device, all heartbeat packets including price tag heartbeat signals received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the electronic price tag system; the server obtains target location information of the product based on the product attributes, and transmits the target location information of the product, all price tag heartbeat signals received by the communication module in the first predetermined time window, and each A readable storage medium having a computer program stored thereon, the computer program being characterized by realizing the steps of: associating the RSSI value at which the heartbeat packet is received with the fingerprint data corresponding to the target location information, and constructing a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained; the communication module transmitting all price tag heartbeat signals received in a second predetermined time window to a server when the movement data of the main body is collected by the motion sensor; and the server matching all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart.

[0065] Those skilled in the art can understand that the implementation of all or part of the flow of the method of the above embodiments can be realized by instructing related hardware by a computer program, and the program can be stored in a non-volatile computer-readable storage medium, and includes the flow of each of the above method embodiments when the program is executed. Here, any reference to memory, storage, database, or other medium used in each embodiment of the present application may include non-volatile memory and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. RAM can come in various forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhancement Type SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Memory Bus (Rambus) Direct RAM (RDRAM), Direct Memory Bus Dynamic RAM (DRDRAM), and Memory Bus Dynamic RAM (RDRAM).

[0066] It should be noted that, in this specification, relational terms such as "first" and "second" are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that such an actual relationship or order exists between those entities or operations. Also, the terms "comprise", "comprising" or any other variation thereof are non-exclusive, and a process, method, article, or device that includes a set of elements not only includes those elements, but also includes other elements not expressly listed or that are inherent to such process, method, article, or device. Unless otherwise specified, an element qualified by the phrase "comprising a" does not exclude the presence of other identical elements in the process, method, article, or equipment that includes the element.

Claims

1. A method for determining a position of a smart shopping cart, comprising: The present invention is applied to a smart shopping cart in a store in which an electronic price tag system including an electronic price tag having known location information and transmitting a heartbeat packet and a server is installed, The smart shopping cart includes a main body, a communication module provided in the main body for receiving a heartbeat packet, a motion sensor for collecting movement data, and a smart device for collecting product attributes; The method for determining the position of the smart shopping cart includes: If product attributes of any product are collected by the smart device, sending all heartbeat packets including a price tag heartbeat signal received by the communication module in a first predetermined time window, an RSSI value at which each heartbeat packet is received, and the product attributes to a server in the EPL system; The server obtains target location information of the product according to the product attributes, associates the target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window and the RSSI value at which each heartbeat packet is received, to obtain fingerprint data corresponding to the target location information, and constructs a positioning fingerprint database corresponding to the store when the server obtains fingerprint data of a predetermined percentage of the store; the communication module transmitting to a server all price tag heartbeat signals received in a second predetermined time window when the body movement data is collected by the motion sensor; the server matches all price tag heartbeat signals received in the second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart; A method for determining a position of a smart shopping cart, comprising:

2. The price tag heartbeat signal includes a price tag ID and a report time; The fingerprint data corresponding to the target location information includes a plurality of price tag IDs and one RSSI weighted average value corresponding to each price tag ID.

2. The method for determining a position of a smart shopping cart according to claim 1.

3. The first predetermined time window is a sum of a time length T1 until the smart device collects the product attributes, a time length T2 from the smart device collecting the product attributes until the shopping cart starts to move, and a time length T3 from the shopping cart starts to move, and the second predetermined time window is a time length T4 until the movement data is collected by the motion sensor.

3. The method for determining a position of a smart shopping cart according to claim 2.

4. The step of the server acquiring target position information of the product based on the product attributes includes: When the server acquires a plurality of pieces of location information based on the product attributes, the server selects the plurality of pieces of location information based on positioning information at a time immediately before the shopping cart is selected, and acquires target location information of the product.

3. The method for determining a position of a smart shopping cart according to claim 2.

5. After constructing a positioning fingerprint database corresponding to the store, As the smart device continues to collect product attributes, acquiring fingerprint data corresponding to a location where the product is located; and updating the positioning fingerprint database with fingerprint data corresponding to the location of the goods.

2. The method for determining a position of a smart shopping cart according to claim 1.

6. Before constructing a positioning fingerprint database corresponding to the store, the server deriving a shelf number and shelf section index corresponding to each EPL based on the price tag IDs in all price tag heartbeat signals received during the second predetermined time window; aggregating and summarizing the price tag heartbeat signals of the same shelf section based on the shelf number and the shelf section index to obtain a summary index corresponding to each shelf section; and determining a target shelf section by comprehensively analyzing the aggregated indicators corresponding to each shelf section, and setting the coordinate position of the target shelf section as positioning information of the smart shopping cart.

2. The method for determining a position of a smart shopping cart according to claim 1.

7. When the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, before constructing a positioning fingerprint database corresponding to the store, calculating an azimuth angle of a emitted signal source of each EPL based on baseband signal characteristics of the predetermined sequence signals received by each antenna; and calculating positioning information of the smart shopping cart based on position information of at least three non-collinear EPL tags and azimuth angles corresponding to the at least three non-collinear EPL tags.

2. The method for determining a position of a smart shopping cart according to claim 1.

8. When the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, the server matches all price tag heartbeat signals received in a second predetermined time window with the positioning fingerprint database to obtain the positioning information of the smart shopping cart, and then: calculating an azimuth angle of a emitted signal source of each EPL based on baseband signal characteristics of the predetermined sequence signals received by each antenna; and further comprising: correcting the positioning information of the smart shopping cart based on the azimuth angle to obtain target positioning information of the smart shopping cart.

2. The method for determining a position of a smart shopping cart according to claim 1.

9. When the communication module includes multiple antennas and the heartbeat packet further includes a predetermined sequence signal, If product attributes of any product are collected by the smart device, transmitting all price tag heartbeat signals received by the communication module in a first predetermined time window, an azimuth angle at which each price tag heartbeat signal is received, and the product attributes to a server in the EPL system; the server acquires location information corresponding to the product based on the product attributes, and acquires location information corresponding to the price tag based on a price tag ID in a price tag heartbeat signal; storing location information corresponding to the price tag and the azimuth angle in an input data set and storing location information corresponding to the item in an output data set; training a machine learning algorithm based on the input dataset and the output dataset to obtain a shopping cart positioning model; and further comprising: inputting price tag position information and an azimuth angle corresponding to the currently received price tag heartbeat signal into a positioning model of the shopping cart for identification by the communication module, thereby obtaining current positioning information of the smart shopping cart.

2. The method for determining a position of a smart shopping cart according to claim 1.

10. A positioning system for a smart shopping cart, comprising: The smart shopping cart includes an electronic price tag whose location information is known, a server, and a smart shopping cart, the smart shopping cart including a main body, a communication module provided in the main body for receiving a heartbeat packet, a motion sensor for collecting movement data, and a smart device for collecting product attributes; the electronic price tag for transmitting a heartbeat packet including a price tag heartbeat signal; the communication module is for, when product attributes of any product are collected by the smart device, transmitting all heartbeat packets received in a first predetermined time window and the product attributes to a server in the EPL system; the server is for obtaining target location information of the product based on the product attributes, associating the target location information of the product with all price tag heartbeat signals received by the communication module in a first predetermined time window as fingerprint data corresponding to the target location information, and constructing a positioning fingerprint database corresponding to the store when a predetermined percentage of fingerprint data of the store is obtained; the communication module is further for transmitting to a server all price tag heartbeat signals received in a second predetermined time window when movement data of the body is collected by the motion sensor; The server is further for matching all price tag heartbeat signals received in a second predetermined time window with the positioning fingerprint database to obtain positioning information of the smart shopping cart. A positioning system for a smart shopping cart.

11. A computing device including a memory, a processor, and a computer program stored in the memory and executable by the processor, Execution of the computer program by the processor implements the steps of the method according to any one of claims 1 to 9.

1. A computer device comprising:

12. A readable storage medium on which a computer program is stored, The computer program is executed by a processor to implement the steps of the method according to any one of claims 1 to 9. A readable storage medium.

Images