Steel cargo pledge financing internet of things supervision method and system based on modification of crown block

Abstract

The application discloses a kind of steel goods pledge financing internet of things supervision method and system based on crane transformation, suitable for the warehouse of goods in and out of warehouse operation by crane.The application is automatically determined by collecting crane hook gravity and the internet of things data of running position, combined with field library area map and crane behavior event determination algorithm, the in and out of warehouse behavior of goods, inventory management is carried out by calculating warehouse inventory change, and early warning information including goods shortage can be sent to financial institution business system.The application has strong universality, high real-time performance, does not need manual participation in supervision operation process, has low dependence on warehouse storage management data, and does not need to change existing management process of warehouse.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of Internet of Things supervision, and particularly relates to a steel cargo pledge financing Internet of Things supervision method and system based on a modification of a crown block. BACKGROUND

[0002] In the commodity pledge financing business of supply chain finance, banks and other financial institutions and warehousing enterprises have begun to cooperate in business in the form of electronic warehouse receipts. In order to reduce business credit risk and ensure that the goods registered in the electronic warehouse receipt are actually stored in the warehouse, the financial institutions and the warehousing enterprises can use the Internet of Things technology to modify the bulk commodity warehouse. On the one hand, the modification improves the registration efficiency of the goods in and out of the warehouse, and on the other hand, the modification uses the Internet of Things supervision measures to prevent the shortage of the goods in the warehouse, which leads to the inconsistency between the total amount of the goods and the electronic warehouse receipt. The commonly used solution in the industry is to use RFID tags to manage the goods one by one. After the RFID tags are pasted on the goods, the operator can register the in and out of the warehouse of the goods through the scanning mode in the warehouse management system (WMS system). The data of the WMS system can be synchronized to the server of the bank and other financial institutions. In addition, a radio frequency machine that senses the RFID tags can be installed at the entrance and exit of the supervised warehouse area, which can sense the behavior of the goods being illegally moved out of the supervised area during the storage period, thereby reducing the financial risk in the electronic warehouse receipt pledge business. However, the existing solution has the following problems:

[0003] 1. The RFID tags communicate with the sensing terminal by means of electromagnetic waves, but the steel and other metal goods have strong electromagnetic shielding properties, which easily causes signal interference problems of the RFID tags pasted on the steel. Once the handheld terminal is far away, the RFID tags cannot be recognized.

[0004] 2. The RFID tags are used to supervise one by one. The steel black material goods are not bound to standardized tags when they are shipped. Therefore, the warehouse management personnel must manually paste the tags when the goods are stored. If the tags need to be reused, the tags also need to be manually removed and recycled when the goods are shipped. This way increases the labor cost and slows down the work efficiency.

[0005] 3. The RFID tags must be pasted on the surface of the goods, but the surface of the steel black material goods may not always be able to paste the RFID tags. For example, the surface of the deformed steel bars (rebar), coiled steel, and angle steel is rugged, and the RFID tags are easily dropped or damaged during the storage process. Therefore, it is difficult to achieve stable binding management of the goods.

[0006] 4. The existing bulk commodity warehouse will also use the data of the overhead crane, the platform scale and the like to check the warehouse-in and warehouse-out operation. However, the basis for the check is that the warehouse party knows in advance the plan of the operation and checks the relevant quality and displacement data. However, the financial institution itself cannot know the operation task of the warehouse party in real time, and the existing conventional check method is not applicable. SUMMARY

[0007] The present application aims at the deficiencies of the prior art and provides a steel cargo pledge financing Internet of Things supervision method and system based on overhead crane modification.

[0008] The present application aims at the deficiencies of the prior art and provides a steel cargo pledge financing Internet of Things supervision method and system based on overhead crane modification.

[0009] (1) Internet of Things modification and data modeling of the storage site

[0010] (1.1) Storage area division and measurement

[0011] The travel range of the overhead crane at the storage site is virtually divided into areas and the areas are measured; it is ensured that the position of any coordinate in the warehouse can uniquely correspond to an virtual area, and a virtual map with coordinates is generated;

[0012] (1.2) Internet of Things data access of the overhead crane

[0013] The quality and displacement Internet of Things data of the hook of the overhead crane are collected; the quality, coordinate and time data are combined into a data structure;

[0014] (1.3) Design of the behavior judgment model of the overhead crane

[0015] The Internet of Things data are analyzed and restored into a behavior event data sequence of the overhead crane, and the corresponding inventory is counted; the Internet of Things data sequence contains the current X-axis coordinate of the overhead crane, the current Y-axis coordinate of the overhead crane, the current quality data and the time corresponding to the current data;

[0016] The behavior event data sequence contains the behavior event type of loading or unloading, the quality of the goods of the current warehouse-in or warehouse-out behavior event and the occurrence time of the behavior event;

[0017] The Internet of Things data sequence is exported into the behavior event data sequence by judging whether the Internet of Things data produce the loading or unloading behavior event and the corresponding quality of the goods, and obtaining the position and time of the occurrence of the behavior event;

[0018] (1.4) Judgment of the warehouse-in and warehouse-out event

[0019] Based on the overhead crane behavior event data sequence, combined with the virtual map, the area where the behavior event occurs is determined, and then it is distinguished whether the event occurs in the supervision area or the non-supervision area; and the overhead crane behavior event data sequence is converted into the overhead crane warehouse entry and exit event sequence, including the corresponding virtual area where the behavior type occurs, the warehouse entry and exit event type, the warehouse entry and exit goods quality, and the occurrence time of the warehouse entry and exit event;

[0020] (2) Internet of Things Warehouse Supervision

[0021] (2.1) Edge computing: upload the corresponding Internet of Things data sequence of the overhead crane behavior event to the cloud platform;

[0022] (2.2) Internet of Things data access: realize the access and subsequent storage of Internet of Things data, and obtain the overhead crane warehouse entry and exit event sequence, that is, each data item represents a warehouse entry and exit event;

[0023] (2.3) Business data integration: obtain electronic warehouse receipts or data related to goods pledge behavior from the business system of financial institutions;

[0024] (2.4) Warehouse management: take each virtual warehouse area as a unit, and maintain the actual inventory of each virtual warehouse area through Internet of Things data and warehouse entry and exit events;

[0025] (2.5) Risk warning: judge whether the warehouse inventory is lower than the weight of the pledged goods, and alarm the relevant person in charge when necessary.

[0026] Further, the entire range of the warehouse overhead crane includes both the supervision area and the non-supervision area, so it is necessary to divide the warehouse into virtual areas and distinguish the displacement of goods loading and unloading through positioning; after the virtual area division, the size of each area is measured to determine the coordinates of each point and the boundary.

[0027] Further, the supervision area needs to be divided again to ensure that each virtual area is rectangular and only stores goods of the same category.

[0028] Further, the Internet of Things data sequence is exported to the behavior event data sequence, which is processed according to the two dimensions of behavior event type and quality, and the specific steps are as follows:

[0029] 1) Behavior event generation determination

[0030] Determine whether loading or unloading behavior events occur in a certain period of time through Internet of Things data; determine whether the quality change in a continuous period of time exceeds the threshold value, if it exceeds the threshold value, it is determined as an effective behavior event;

[0031] 2) Behavior event type determination

[0032] If the IoT data sequence in a certain period of time is identified as a valid behavior event, it is necessary to determine whether it corresponds to loading or unloading, i.e., to determine the type of behavior event; the quality sequence shows an overall downward trend over time, which is considered to be unloading behavior; otherwise, it is considered to be loading behavior;

[0033] 3) Quality determination

[0034] If the IoT data sequence is identified as a loading behavior, the corresponding cargo quality needs to be estimated; if it is unloading behavior, it is directly recognized as empty;

[0035] 4) Behavior event location determination

[0036] If the original IoT data sequence is identified as a valid behavior event, the location of the behavior event needs to be determined to confirm the corresponding virtual area; the average value of the X-axis and Y-axis coordinates is determined as the coordinates of the behavior event;

[0037] 5) Behavior event time determination

[0038] If the IoT data sequence is identified as a valid behavior event, the median of each timestamp in the sequence is taken as the occurrence time of the behavior event;

[0039] Finally, the IoT data sequence is converted into a behavior event data sequence through the above rules.

[0040] Further, based on the obtained crane behavior event data sequence, two types of crane event behaviors can be obtained: loading and unloading; combined with the virtual map, the occurrence location of the two types of crane event behaviors can be further obtained, i.e., whether it is outside or inside the warehouse, thereby obtaining four types of corresponding warehouse entry and exit events, as follows:

[0041] a. From the loading behavior outside the warehouse to the unloading behavior inside the warehouse, the corresponding warehouse entry and exit event is warehouse entry;

[0042] b. From the loading behavior inside the warehouse to the unloading behavior outside the warehouse, the corresponding warehouse entry and exit event is warehouse exit;

[0043] c. From the loading behavior outside the warehouse to the unloading behavior outside the warehouse, the corresponding warehouse entry and exit event is warehouse movement outside, and this warehouse entry and exit event is not counted;

[0044] d. From the loading behavior inside the warehouse to the unloading behavior inside the warehouse, the corresponding warehouse entry and exit event is warehouse movement.

[0045] Further, in step (1.2), if a gravity sensor is integrated on the crane, the sensor data is accessed through the edge gateway, and the quality data is accessed to the cloud; if the gravity sensor is not integrated on the crane, an electronic sling scale needs to be installed under the crane hook, and the electronic sling scale communicates with the edge gateway in a wireless manner and accesses the quality data to the cloud.

[0046] Further, in step (1.2), if the overhead crane is integrated with a displacement sensor, access the sensor data through the edge gateway and upload the displacement data to the cloud; if the overhead crane is not integrated with a displacement sensor, install proximity switches on the overhead crane track at intervals of 50 cm and record the coordinates of each proximity switch; when the overhead crane passes, the proximity switch is triggered, indicating that the overhead crane is at the coordinate of the proximity switch.

[0047] Further, the warehouse management includes a warehouse updating and a warehouse calibration process;

[0048] Warehouse updating: through the sequence of overhead crane in and out of the warehouse events, dynamically adjust the actual inventory of each warehouse area;

[0049] Warehouse calibration: warehouse updating only provides the change amount of each area inventory, and needs to set an initial value for each area inventory through warehouse calibration; at the same time, considering that the Internet of Things data will produce accumulated errors in the actual process, calibration needs to be carried out regularly, and the data source of warehouse calibration is the artificial inventory data provided by the warehouse party.

[0050] The application also provides a steel cargo pledge financing Internet of Things supervision system based on overhead crane transformation, which comprises an Internet of Things transformation and data modeling module and an Internet of Things warehouse supervision module in a warehouse site;

[0051] The warehouse site Internet of Things transformation and data modeling module comprises a warehouse area division and measurement sub-module, an overhead crane Internet of Things data acquisition sub-module, an overhead crane behavior judgment model design sub-module and an in and out of warehouse event judgment sub-module

[0052] The warehouse area division and measurement sub-module is used for virtually dividing the travel range of the overhead crane in the warehouse site and measuring the area; ensuring that the position of any coordinate in the warehouse can be uniquely corresponding to a virtual area, and generating a virtual map with coordinates;

[0053] The overhead crane Internet of Things data acquisition sub-module is used for collecting the hook quality and displacement Internet of Things data of the overhead crane; combining the quality, coordinate and time data into a data structure;

[0054] The crane behavior judgment model design submodule is used for analyzing the Internet of Things data, restoring the Internet of Things data into a crane behavior event data sequence, and counting corresponding inventory; the Internet of Things data sequence comprises a current crane X-axis coordinate, a current crane Y-axis coordinate, current quality data, and a current data corresponding time; the behavior event data sequence comprises a loading or unloading behavior event type, a cargo quality of the current loading or unloading behavior event, and a time of occurrence of the current behavior event; the Internet of Things data sequence is exported into the behavior event data sequence by judging whether the Internet of Things data generates a loading or unloading behavior event and corresponding cargo quality, and obtaining a behavior event occurrence position and time;

[0055] The warehouse in-out event judgment submodule is used for judging a region where a behavior event occurs based on the crane behavior event data sequence and a virtual map, and further distinguishing whether the event occurs in a supervision area or a non-supervision area; and the crane behavior event data sequence is converted into a crane warehouse in-out event sequence, comprising a corresponding virtual area where a behavior type occurs, a current warehouse in-out event type, a cargo quality of the current warehouse in-out, and a time of occurrence of the current warehouse in-out event.

[0056] The Internet of Things warehouse supervision module comprises an edge computing submodule, an Internet of Things data access submodule, a business data integration submodule, a warehouse management submodule, and a risk early warning submodule.

[0057] The edge computing submodule is used for uploading a corresponding Internet of Things data sequence generating a crane behavior event to a cloud platform.

[0058] The Internet of Things data access submodule is used for realizing access and subsequent storage of Internet of Things data, and obtaining a crane warehouse in-out event sequence, that is, each data item represents a cargo warehouse in-out event.

[0059] The business data integration submodule is used for obtaining data related to electronic warehouse receipts or cargo pledge behaviors from a business system of a financial institution.

[0060] The warehouse management submodule is used for maintaining actual inventory of each virtual warehouse area through the Internet of Things data and the cargo warehouse in-out event, with each virtual warehouse area as a unit.

[0061] The risk early warning submodule is used for judging whether warehouse inventory is lower than the weight of pledged cargo, and alarming relevant persons in charge when necessary.

[0062] Advantages of the present application:

[0063] 1. The present application does not need to paste RFID on the cargo, and when the warehouse is transformed into an Internet of Things, it does not need to consider arranging a person to paste and remove the label, and does not increase labor costs.

[0064] 2. This invention does not require RFID tags to be affixed, and there is no need to consider the problem of tags falling off or being damaged during warehouse operations.

[0065] 3. This invention can directly determine inbound and outbound operations using the behavior of the overhead crane, and does not strictly require verification by connecting to the warehouse management system. Attached Figure Description

[0066] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0067] Figure 1 A flowchart of an IoT-based monitoring method for steel cargo financing based on overhead crane modification, provided for this invention;

[0068] Figure 2 A schematic diagram illustrating an example of a virtual region partitioning method based on this invention for any regulatory warehouse area;

[0069] Figure 3 This is a flowchart illustrating the use of IoT data to determine the behavior of overhead cranes according to the present invention.

[0070] Figure 4 This invention provides an application architecture diagram for an IoT system for steel cargo financing based on overhead crane modification. Detailed Implementation

[0071] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0072] like Figure 1 As shown, this invention provides banks and other financial institutions with an IoT-based monitoring method for steel cargo-backed financing, based on overhead crane modifications. It is applicable to warehouses where goods are handled via overhead cranes for inbound and outbound operations. When conducting bulk commodity-backed financing, financial institutions primarily calculate the value of the collateral based on the total mass of the goods. This invention modifies the overhead crane to collect relevant mass and displacement data, detecting the mass and displacement of goods entering and leaving the warehouse, thus achieving monitoring of the total mass of goods in the supervised warehouse area. This invention does not rely on RFID tag technology, therefore eliminating the need for manual RFID tag affixing or removal from goods, without affecting the warehouse's operational processes, and without concerns about tag detachment.

[0073] Generally, the warehouse itself relies on the crane to weigh the goods in and out of the warehouse, but this behavior usually needs to know the occurrence of the event in advance and be verified by Internet of Things data. For example, the operator registers the in and out operation in the warehouse management system, and the behavior is evidenced by the crane data. However, in the financial institution's pledge financing supervision scene, the financial institution itself cannot know when and where the warehouse needs to perform a certain operation, and the crane data cannot be directly used. Generally, the financial institution and the warehouse need to interface with the warehouse management system data. The supervision method described in the present application maximizes the reduction of the interface requirement of the financial institution and the warehouse management system of the warehouse on the basis of using the crane data, and has strong applicability.

[0074] The present application automatically determines the in and out behavior of the goods by collecting the Internet of Things data of the crane hook gravity and the running position, combining the on-site warehouse area map and the crane behavior event determination algorithm, managing the warehouse inventory by calculating the warehouse inventory change, and sending warning information including goods shortage to the financial institution business system. The present application has strong universality, high real-time performance, does not require manual participation in the supervision operation process, has low dependence on the warehouse management system data of the warehouse, and does not need to change the existing management process of the warehouse.

[0075] The present application provides a steel pledge financing Internet of Things supervision method based on crane modification, and the specific steps are as follows:

[0076] (1) Internet of Things modification and data modeling of the warehouse site

[0077] (1.1) Warehouse area division and measurement

[0078] The travel range of the crane at the warehouse site is virtually divided into areas and the areas are measured. This scheme relies on positioning information to determine the in and out behavior of the goods. Since the entire travel range of the crane in the warehouse includes the supervision area (for example, inside the warehouse) and the non-supervision area (for example, outside the warehouse), it is necessary to virtually divide the warehouse into areas in order to distinguish the displacement of the goods loading and unloading by positioning. In addition, since the supervision area is not necessarily a simple geometric shape (for example, a rectangle), or there may be multiple types of goods (for example, disc screws and threaded steel) stored in the supervision area, the supervision area is generally divided again to finally ensure that each virtual area is a rectangle and only stores the same type of goods. After virtual area division, the size of each area is measured to determine the coordinates of each point and the boundary, and a virtual map with coordinates is generated.

[0079] In actual scenarios, the warehouse can be divided into any number of rectangular virtual areas to ensure that any coordinate position in the warehouse can be uniquely matched to a virtual area. For ease of understanding, the virtual area division in the example of the present application is shown in Figure 2 , in which virtual area A is a non-supervision area (i.e., outside the warehouse), and virtual areas B and C are different supervision areas.

[0080] (1.2) Obtain the overhead crane Internet of Things data

[0081] Collect the hook of the overhead crane Collect quality and displacement Internet of Things data;

[0082] 1) Quality collection: if the overhead crane is integrated with a gravity sensor, the sensor data can be accessed through the edge gateway, and the quality data is accessed to the cloud; if the overhead crane is not integrated with a gravity sensor, an electronic sling scale needs to be installed under the overhead crane hook, and the electronic sling scale communicates with the edge gateway in a wireless manner, and the quality data is accessed to the cloud. For convenience, define the quality data Wm, unit ton.

[0083] 2) Displacement collection: if the overhead crane is integrated with a displacement sensor, the sensor data can be accessed through the edge gateway, and the displacement data is accessed to the cloud; if the overhead crane is not integrated with a displacement sensor, proximity switches can be installed every 50 centimeters on the overhead crane track, and the coordinates of each proximity switch are recorded. When the overhead crane passes, the proximity switch is triggered, indicating that the overhead crane is at the coordinate of the proximity switch. The displacement of the general overhead crane is a two-dimensional plane, and for convenience, a plane index coordinate system XoY is defined on the overhead crane running platform, the X-axis coordinate is Xm, and the Y-axis is Ym, both units are meters.

[0084] 3) Considering the data transmission scale and data continuity requirements, the typical value of the Internet of Things data collection interval is 500 milliseconds. In order to apply the subsequent data, the quality, coordinate and time data need to be combined into a data structure. Let the original Internet of Things data structure be Am, the data structure format of Am is {Xm, Ym, Wm, Tm}, Tm is the data collection time (unit: milliseconds), Wm is the collected quality at Tm, Xm and Ym are the X-axis and Y-axis coordinates of the overhead crane running position at Tm.

[0085] (1.3) Design of overhead crane behavior judgment model

[0086] Since the user of the present application is a financial institution. The financial institution usually cannot assign a special person to continuously supervise the warehouse operation behavior in the warehouse site, and the quality and displacement original Internet of Things data collected in the overhead crane modification cannot directly reflect the actual behavior of the overhead crane (i.e. loading or unloading), therefore a mathematical model needs to be established to analyze the original Internet of Things data and restore the overhead crane behavior event data sequence, so as to facilitate the financial institution to count the corresponding inventory accordingly.

[0087] Let the original Internet of Things data sequence be {Am}, the data structure format of a certain item (i.e. the mth item) Am in the sequence is {Xm, Ym, Wm, Tm}. Among them, Xm is the current X-axis coordinate of the overhead crane, Ym is the current Y-axis coordinate of the overhead crane, Wm is the current quality data (unit: ton), and Tm is the time corresponding to the current data (i.e. time stamp).

[0088] Let the sequence of behavioral events be {Bn}, and the data format of a certain item (i.e., the nth item) An in the sequence be {Hn, Xn, Yn, Wn, Tn}. Here, Hn is the type of behavioral event, which includes two meaningful cases: loading (i.e., loading goods from an empty state) and unloading (i.e., unloading loaded goods to restore an empty state); Wn is the quality value of the goods in this outbound or inbound behavioral event (in tons); and Tn is the time when this behavioral event occurred.

[0089] We need to provide a model to derive a behavioral event data sequence {Bn} from the raw IoT data sequence {An}, and process it separately according to two dimensions: behavioral event type Hn and quality Wn. For example... Figure 3 As shown, the specific steps are as follows:

[0090] 1) Determination of the occurrence of behavioral events

[0091] The Internet of Things (IoT) data is used to determine whether loading or unloading events occurred within a certain period. This can be determined by whether the change in quality value within a finite time exceeds a threshold.

[0092] Take the original IoT data sequence {A m The i-th term (A) in} i Starting with k consecutive terms (k is a model parameter that needs to be manually adjusted), a subsequence {Ak} is formed. m}={A i A i+1 A i+2 ,...,A k-i+1}. For the subsequence {Ak} m For each item in}, extract its current quality data and generate the corresponding quality subsequence {Wm}. k}={W i W i+1 W i+2 ,...,W i-m+1}. The mass subsequence {Wk} is obtained. m The maximum value in} is Wmax, the minimum value is Wmin, and the quality change threshold Ws is defined (Ws is a model parameter that needs to be manually adjusted and determined). If |Wmax-Wmin|>Ws, it can be considered that within k terms (i.e., within a finite time), the quality value change exceeds the threshold, and this can be considered a valid behavioral event, i.e., loading or unloading.

[0093] 2) Determination of Behavioral Event Type Hn

[0094] If the original IoT data subsequence {Ak} within a certain time period m If a behavior event is identified as a valid behavior event, it is necessary to determine whether it corresponds to loading or unloading, that is, to determine the type of behavior event.

[0095] Take {Ak m} Corresponding mass subsequence {Wk m}, define the data percentile p (p is a percentage number, ranging from 0-100%, which needs to be manually determined), and set {Wk m Summing the first p terms and the last p terms of} yields Sp1 and Sp2 respectively (e.g., {Wk} m There are 100 items in total. If p is 5%, then Sp1 is the sum of its first 5 items and Sp2 is the sum of its last 5 items. If Sp1 > Sp2, the quality sequence shows an overall downward trend over time, which can be considered as unloading behavior; conversely, if Sp1 > Sp2, the quality sequence shows an overall upward trend over time, which can be considered as loading behavior.

[0096] 3) Quality Wn Judgment

[0097] If the original IoT data subsequence {Ak} m If a loading action is identified as a loading action, the corresponding cargo mass needs to be estimated; if it is an unloading action, the corresponding Wn can be directly considered to be 0, indicating an empty load.

[0098] According to Newton's second law, positive and negative accelerations exist during the loading or unloading of goods, resulting in {Ak} m The corresponding mass subsequence {Wk} m In the equation, some items are greater than the actual mass, and some items are less than the actual mass. The actual mass Wm needs to be estimated using the following method:

[0099] a) For the mass subsequence {Wk} m Sort the subsequences by size and remove all quality values ​​that are less than the first quartile and greater than the third quartile (to remove values ​​that deviate significantly). Denote the processed subsequence as {Wk'}. m}

[0100] b) For {Wk' m Each item Wk' in} i Find it and the previous term (Wk') respectively. i-1 ) and the next term (Wk' i+1 Find the difference between the two differences, and sum the two differences to obtain the corresponding sum of differences Dk. i .

[0101] c) Filter out Dk i The minimum value, corresponding to Wk' i This can be considered as the mass value at a temporary steady state during the lifting process, and it is taken as the cargo mass W corresponding to this loading event. n The estimated value.

[0102] 4) Location X of the behavioral eventn with Y n determination

[0103] If the original IoT data subsequence {Ak m} is identified as a valid behavior event, the position of the behavior event needs to be determined in order to confirm the corresponding virtual area.

[0104] According to the crane operation specification, the crane is not allowed to move horizontally (i.e., not allowed to move in the X-axis and Y-axis directions) during loading and unloading. It can be considered that the X n with Y n will not change significantly during the occurrence of a behavior event. Therefore, the average value of the X-axis coordinate of each item in the {Ak m} subsequence can be determined as the coordinate X n of the occurrence of the behavior event, and the average value of the Y-axis coordinate of each item can be determined as the coordinate Y n of the occurrence of the behavior event. This approximation can meet the application requirements.

[0105] 5) Behavior event occurrence time T n determination

[0106] If the original IoT data subsequence {Ak m} is identified as a valid behavior event, the median of the time stamp of each item in the {Ak m} subsequence is taken as the occurrence time T n of the behavior event. Since this model does not have high accuracy requirements for the occurrence time of the behavior event, this approximation can meet the application requirements and does not require additional computational complexity.

[0107] 6) De-duplication principle

[0108] Through the above method, the IoT data subsequence {Ak m} that meets the conditions is determined as a behavior event B n , and the corresponding {H n , W n , T n} are determined. When selecting the next group of IoT data subsequences, all items in {Ak m} will be skipped, i.e., starting from the i+kth item, which ensures that the next subsequence does not contain any item in the previous subsequence {Ak m} to prevent the same group of data from being repeatedly determined. If the IoT data subsequence {Ak m} does not correspond to a behavior event, the next subsequence starts from the i+1th item (i.e., {A i+1 , A i+2 , A i+3 ,..., A k-i}) and continues to be determined.

[0109] Finally, the original IoT data sequence {A m} can be transformed into the behavior event data sequence {B n} by the above rules. And used for subsequent warehouse entry and exit event determination.

[0110] (1.4) Parameter debugging

[0111] In the above described crane behavior determination model, there are several parameters that need to be set artificially, as follows:

[0112] Parameter k: the number of items of the IoT original data subsequence {Ak m}, if set too large, it may lead to two behaviors contained in a sequence, if too small, it may lead to that the subsequence cannot feedback the whole process of a behavior event.

[0113] Parameter Ws: the quality change threshold Ws, used to define the sensitivity of the crane to the quality change, generally set according to the typical warehouse goods quality, if set too large, it may lead to some cargo loading and unloading behavior being overlooked, if too small, it may lead to some noise data being identified as a behavior event.

[0114] Parameter p: the percentage of the quality subsequence for change trend judgment. The determination of the event behavior is based on the calculation of the overall upward or downward trend of the quality subsequence over time, and the parameter p represents the window size of the sampling data. If set too small, noise may be considered as a typical quality value; if set too large, data may be affected by the quality value of the acceleration during loading or unloading.

[0115] The above parameters need to be compared according to the actual scene, so that the model can feedback the real situation.

[0116] In addition, based on the virtual map divided in step (1.1), the crane driving position needs to be calibrated, so that the two can be in the same reference system and can be directly compared. Finally, taking a set of event behavior data coordinates X n and Y n , the virtual area where the behavior occurs can be determined.

[0117] (1.5) Warehouse entry and exit event determination

[0118] Based on the above crane behavior determination model, the crane behavior event data sequence {B n} can be obtained, combined with the virtual map, any item B n in B iThe corresponding behavior event occurs in the area, and then distinguishes whether the event occurs in the supervision area or the non-supervision area. For the convenience of description, it is assumed that the warehouse is divided into virtual areas A, B and C. It is assumed that area A is a non-supervision area (i.e. outside the warehouse), and areas B and C are different supervision areas.

[0119] Therefore, for any one overhead crane behavior event B i , there are the following four cases: loading outside the warehouse, unloading outside the warehouse, loading inside the warehouse, and unloading inside the warehouse. Since one specific warehouse entry and exit event of the overhead crane must exist in a pair of behavior events (for example, the behavior event corresponding to the warehouse entry event is the combination of loading outside the warehouse and unloading inside the warehouse, which are referred to as behavior 1 and behavior 2, respectively, below), a mechanism needs to be established to further convert the overhead crane behavior event data sequence {B n} into an overhead crane warehouse entry and exit event sequence, denoted as {C p}.

[0120] The data structure format of a certain item (i.e. the i-th item) C p in the sequence {C i} is {L1 i , L2 i , H i , W i , T i}, where L1 i represents the virtual area where behavior 1 occurs, L2 i represents the virtual area where behavior 2 occurs, H i indicates the type of warehouse entry and exit event (only including warehouse entry, warehouse exit, and warehouse transfer), W i indicates the weight of the goods in this warehouse entry and exit, and T i indicates the occurrence time of this warehouse entry and exit event.

[0121] In theory, the overhead crane should not load twice in succession or unload twice in succession. Four kinds of legal overhead crane warehouse entry and exit event sequences are listed in the form of enumeration as shown in Table 1:

[0122] Table 1

[0123] Action 1 Action 2 Corresponding entry and exit warehouse events Loading outside warehouse Unloading inside warehouse Entry Loading inside warehouse Unloading outside warehouse Exit Loading outside warehouse Unloading outside warehouse Moving outside warehouse (not counted) Loading inside warehouse Unloading inside warehouse Moving warehouse

[0124] It can be seen that a legal behavior must start with a loading event and end with an unloading event. Now, for the obtained overhead crane behavior event data sequence {B n}, the first item B1 starts to traverse the sequence and performs the following processing:

[0125] a) Take any item B i and determine whether B i corresponds to a loading behavior and B i+1 corresponds to an unloading behavior. If so, proceed to the next step, otherwise take the next item B i+1Rejudge.

[0126] b) If B i corresponds to the behavior of loading in the warehouse, B i+1 corresponds to the behavior of unloading outside the warehouse, it is considered that B i and B i+1 constitute a warehouse-in event. If the crane in and out of the warehouse event is counted as C i , then the corresponding L1 i is B i , the virtual area (for example, A) where L2 i occurs is B i+1 , the virtual area (for example, B) where H i occurs is “warehouse-in”, W i is B i , the weight corresponding to the behavior event T i is B i+1 , and the timestamp corresponding to the behavior event is B.

[0127] c) If B i corresponds to the behavior of loading in the warehouse, B i+1 corresponds to the behavior of unloading outside the warehouse, it is considered that B i and B i+1 constitute a warehouse-out event. If the crane in and out of the warehouse event is counted as C i , then the corresponding L1 i is B i , the virtual area (for example, B) where L2 i occurs is B i+1 , the virtual area (for example, A) where H i occurs is “warehouse-out”, W i is B i , the weight corresponding to the behavior event T i is B i+1 , and the timestamp corresponding to the behavior event is B.

[0128] d) If B i corresponds to the behavior of loading in the warehouse, B i+1 corresponds to the behavior of unloading in the warehouse, it is considered that B i and B i+1 constitute a warehouse-moving event. If the crane in and out of the warehouse event is counted as C i , then the corresponding L1 i is B i , the virtual area (for example, B) where L2 i occurs is B i+1 , the virtual area (for example, C) where H i occurs is “warehouse-moving”, W i is B i , the weight corresponding to the behavior event T i is B i+1 , and the timestamp corresponding to the behavior event is B.

[0129] e) For other cases (intra-warehouse movement or inter-warehouse movement) are not counted, take the next item B i+1 Re-determine.

[0130] Finally, the model can be obtained through the crane behavior event data sequence {B n} , the crane in and out of the warehouse event sequence {C p}, and the Internet of Things warehouse supervision for subsequent use.

[0131] (2) Internet of Things Warehouse Supervision

[0132] The Internet of Things warehouse supervision is distributed in the business site (that is, the place where the actual business occurs, generally a bulk commodity warehouse that has been Internet of Things transformed) and the business background (that is, a financial institution or a technology company that provides related supervision services, generally deployed in a server room). Specifically, the following steps are included:

[0133] (2.1) Edge Computing

[0134] Considering that the amount of original Internet of Things data is large, and there are multi-dimensional data (including quality, displacement, and other physical quantities), in order to reduce data transmission load and improve data processing efficiency, edge computing is needed. Its function is to collect original Internet of Things data to form a data subsequence {A m}, determine the subsequence {Ak m} in {A m}, and only upload the data subsequence {Ak m} that produces behavior events to the cloud platform, as shown in Figure 4 .

[0135] (2.2) Internet of Things Data Access

[0136] Internet of Things data access and subsequent storage are realized, and in this invention, it also needs to bear the subsequent part of the crane behavior determination model, and finally obtain the crane in and out of the warehouse event sequence {Cp}, that is, each data item represents a cargo in and out of the warehouse event.

[0137] (2.3) Business Data Integration

[0138] The application of Internet of Things data is ultimately for the service of cargo pledge financing business. Through business data integration, electronic warehouse receipts or other data related to cargo pledge behavior are obtained from the business system of financial institutions. The fundamental role of business data is to provide the total weight basis of each warehouse area pledged cargo, so as to compare the actual cargo weight reflected by the Internet of Things data with the pledged cargo weight later.

[0139] (2.4) Warehouse Management

[0140] The actual inventory of each virtual warehouse area is maintained through the Internet of Things data in units of each virtual warehouse area. The warehouse management provides warehouse updating and warehouse calibration functions.

[0141] Warehouse updating: that is, through the crane in and out of the warehouse event sequence {C p}, the actual inventory of each warehouse area is dynamically adjusted. For a specific C i , the data structure format is {L1 i , L2 i , H i , W i , T i}, the inventory updating is performed through the following strategy. It is assumed that a warehouse is divided into virtual areas A, B and C, where area A represents a non-supervised area (for example, outside the warehouse), and areas B and C represent different supervised areas. As shown in Table 2, events that can occur in C i are shown in Table 2, and the corresponding warehouse updating behaviors are shown in Table 2. (B and C are completely equivalent, and only the following three possibilities are included after removing the duplicate cases).

[0142] Table 2

[0143]

[0144]

[0145] Warehouse calibration: the warehouse updating function only provides the change amount of the inventory of each area, and the initial value of the inventory of each area needs to be set through the warehouse calibration function. At the same time, considering that the Internet of Things data will accumulate errors in the actual process, the function needs to be calibrated regularly to prevent the deviation between the Internet of Things data and the actual data from being too large. The data source of the warehouse calibration can be manual inventory data provided by the warehouse party.

[0146] (2.5) Risk warning

[0147] A risk warning function is provided. In the present application, the most typical risk warning is that the warehouse inventory is less than the weight of the pledged goods, at which time the system can be connected, short message, email form to the relevant person in charge of warning. Users can design other specific risk rules according to actual conditions.

[0148] On the other hand, the present application also provides a steel cargo pledge financing Internet of Things supervision system based on crane modification, which comprises an Internet of Things modification and data modeling module and an Internet of Things warehouse supervision module in the warehouse site; the specific implementation process of each module is described with reference to the implementation steps of the steel cargo pledge financing Internet of Things supervision method based on crane modification provided by the present application.

[0149] The Internet of Things transformation and data modeling module of the warehouse site comprises a warehouse area division and measurement sub-module, an overhead traveling crane Internet of Things data acquisition sub-module, an overhead traveling crane behavior judgment model design sub-module, and an in-out warehouse event judgment sub-module

[0150] The warehouse area division and measurement sub-module is used for virtually dividing the travel range of the overhead traveling crane in the warehouse site and performing area measurement; ensuring that the position of any coordinate in the warehouse can correspond to a virtual area uniquely, and generating a virtual map with coordinates;

[0151] The overhead traveling crane Internet of Things data acquisition sub-module is used for collecting the hook collection quality and displacement Internet of Things data of the overhead traveling crane; and combining the quality, coordinate, and time data into a data structure;

[0152] The overhead traveling crane behavior judgment model design sub-module is used for analyzing the Internet of Things data, restoring the Internet of Things data sequence into overhead traveling crane behavior event data, and counting the corresponding inventory; the Internet of Things data sequence comprises the current X-axis coordinate of the overhead traveling crane, the current Y-axis coordinate of the overhead traveling crane, the current quality data, and the corresponding time data; the behavior event data sequence comprises the behavior event type of loading or unloading, the cargo quality of the in-out warehouse behavior event, and the occurrence time of the behavior event; the Internet of Things data sequence is exported into the behavior event data sequence by judging whether the Internet of Things data produces the loading or unloading behavior event and the corresponding cargo quality, and obtaining the behavior event occurrence position and time;

[0153] The in-out warehouse event judgment sub-module is used for judging the area where the behavior event occurs based on the overhead traveling crane behavior event data sequence and the virtual map, and further distinguishing whether the event occurs in the regulated area or the non-regulated area; and converting the overhead traveling crane behavior event data sequence into the overhead traveling crane in-out warehouse event sequence, comprising the corresponding virtual area where the behavior type occurs, the in-out warehouse event type, the cargo quality of the in-out warehouse, and the occurrence time of the in-out warehouse event;

[0154] The Internet of Things warehouse supervision module comprises an edge computing sub-module, an Internet of Things data access sub-module, a business data integration sub-module, a warehouse management sub-module, and a risk early warning sub-module;

[0155] The edge computing sub-module is used for uploading the corresponding Internet of Things data sequence of the overhead traveling crane behavior event to the cloud platform;

[0156] The Internet of Things data access sub-module is used for realizing the access and subsequent storage of the Internet of Things data, and obtaining the overhead traveling crane in-out warehouse event sequence, that is, each data item represents an in-out warehouse event of a cargo;

[0157] The business data integration sub-module is used for obtaining the data related to the electronic warehouse receipt or the cargo pledge behavior from the business system of the financial institution;

[0158] The warehouse management submodule is used for maintaining actual inventory of each virtual warehouse area through Internet of Things data and cargo warehouse-in and warehouse-out events respectively in units of each virtual warehouse area;

[0159] The risk early warning submodule is used for judging whether the warehouse inventory is lower than the weight of the pledged cargo, and alarming the relevant person in charge when necessary.

[0160] The above examples are used to explain and illustrate the present application, but not to limit the present application, any modification and change made to the present application within the spirit and protection scope of the claims of the present application, falls into the protection scope of the present application.

Claims

1. A steel cargo pledge financing Internet of Things supervision method based on a modification of a crown block, characterized by, The specific steps of the method are as follows: (1) Warehouse site Internet of Things transformation and data modeling (1.1) Warehouse area division and measurement The entire travel range of the warehouse overhead crane includes both the regulated area and the non-regulated area, so it is necessary to virtually divide the warehouse area and distinguish the displacement of cargo loading and unloading through positioning; after virtual area division, the size of each area is measured to determine the coordinates of each point and the boundary; ensure that any coordinate position in the warehouse can uniquely correspond to a virtual area, and generate a virtual map with coordinates; (1.2) Obtain overhead crane Internet of Things data Collect the overhead crane hook quality and displacement Internet of Things data; combine quality, coordinate, and time data into a data structure; (1.3) Overhead crane behavior judgment model design Analyze the Internet of Things data and restore it to a crane behavior event data sequence, and count the corresponding inventory; the Internet of Things data sequence includes the current X-axis coordinate of the crane, the current Y-axis coordinate of the crane, the current quality data, and the current data corresponding time; The behavior event data sequence includes the behavior event type of loading or unloading, the cargo quality of this time out-of-warehouse or in-warehouse behavior event, and the occurrence time of this behavior event; By judging whether the Internet of Things data produces a loading or unloading behavior event and the corresponding cargo quality, the behavior event occurrence position and time are obtained, and the Internet of Things data sequence is exported to the behavior event data sequence; (1.4) Out-of-warehouse or in-warehouse event judgment Based on the crane behavior event data sequence, combined with the virtual map, the area where the behavior event occurs is judged, and then it is distinguished whether the event occurs in the regulated area or the non-regulated area; and the crane behavior event data sequence is converted into the crane out-of-warehouse or in-warehouse event sequence, including the corresponding virtual area where the behavior type occurs, the type of this time out-of-warehouse or in-warehouse event, the cargo quality of this time out-of-warehouse or in-warehouse, and the occurrence time of this time out-of-warehouse or in-warehouse event; the specific implementation is as follows: 1) Behavior event generation judgment Determine whether a loading or unloading behavior event occurs in a certain period of time through Internet of Things data; whether the quality changes in a continuous period of time exceeds the threshold value is determined, if it exceeds the threshold value, it is determined as an effective behavior event; 2) Behavior event type judgment If the Internet of Things data sequence in a certain period of time is identified as an effective behavior event, it is necessary to determine whether it corresponds to loading or unloading, that is, to judge the type of behavior event; if the quality sequence shows a general downward trend over time, it is considered to be an unloading behavior; otherwise, it is considered to be a loading behavior; 3) Quality judgment If the Internet of Things data sequence is identified as a loading behavior, the corresponding cargo quality needs to be estimated; If it is an unloading behavior, it is directly recognized as empty; 4) Behavior event occurrence position judgment If the original Internet of Things data sequence is identified as an effective behavior event, the position where the behavior event occurs needs to be determined to confirm the corresponding virtual area; the average value of the X-axis and Y-axis coordinates is determined as the coordinate of the occurrence of this behavior event; 5) Behavior event occurrence time judgment If the Internet of Things data sequence is identified as an effective behavior event, the median of each time stamp in the sequence is taken as the occurrence time of this behavior event; Finally, the IoT data sequence is converted into the behavior event data sequence by the above rules; (2) IoT warehouse supervision (2.1) Edge computing: upload the corresponding IoT data sequence of the crane behavior event to the cloud platform; (2.2) IoT data access: realize the access and subsequent storage of IoT data, and obtain the crane in-and-out warehouse event sequence, that is, each data item represents a cargo in-and-out warehouse event; (2.3) Business data integration: obtain electronic warehouse receipts or data related to cargo pledge behavior from the business system of financial institutions; (2.4) Warehouse management: take each virtual warehouse area as a unit, and maintain the actual inventory of each virtual warehouse area through IoT data and cargo in-and-out warehouse events; (2.5) Risk warning: judge whether the warehouse inventory is lower than the weight of the pledged goods, and alarm the relevant person in charge when necessary.

2. The method according to claim 1, wherein, The supervision area needs to be divided again to ensure that each virtual area is rectangular and only stores the same type of goods.

3. The method according to claim 1, wherein the method is based on a crane- modified steel cargo pledge financing Internet of Things supervision method. Based on the obtained crane behavior event data sequence, two crane event behaviors can be obtained: loading and unloading; combined with the virtual map, the occurrence position of the two crane event behaviors can be further obtained: inside or outside the warehouse, thereby obtaining four kinds of corresponding in-and-out warehouse events, as follows: a. From the behavior of loading outside the warehouse to the behavior of unloading inside the warehouse, the corresponding in-and-out warehouse event is warehouse entry; b. From the behavior of loading inside the warehouse to the behavior of unloading outside the warehouse, the corresponding in-and-out warehouse event is warehouse exit; c. From the behavior of loading outside the warehouse to the behavior of unloading outside the warehouse, the corresponding in-and-out warehouse event is warehouse movement outside; d. From the behavior of loading inside the warehouse to the behavior of unloading inside the warehouse, the corresponding in-and-out warehouse event is warehouse movement.

4. The method according to claim 1, wherein, In step (1.2), if the crane is integrated with a gravity sensor, access the sensor data through the edge gateway, and upload the quality data to the cloud; if the crane is not integrated with a gravity sensor, an electronic hoist scale needs to be installed under the crane hook, and the electronic hoist scale communicates with the edge gateway in a wireless manner and uploads the quality data to the cloud.

5. The method according to claim 1, wherein, In step (1.2), if the crane is integrated with a displacement sensor, access the sensor data through the edge gateway, and upload the displacement data to the cloud; if the crane is not integrated with a displacement sensor, install proximity switches every 50 centimeters on the crane track, and record the coordinates of each proximity switch; when the crane passes, the proximity switch is triggered, indicating that the crane is at the coordinate of the proximity switch.

6. The method according to claim 1, wherein, Warehouse management includes warehouse updating and warehouse calibration processes; Warehouse updating: dynamically adjust the actual inventory of each warehouse area through the crane in-and-out warehouse event sequence; Warehouse calibration: warehouse updating only provides the change amount of the inventory of each area, and needs to set an initial value for the inventory of each area through warehouse calibration; at the same time, considering that the IoT data will accumulate errors in the actual process, calibration needs to be performed regularly, and the data source of warehouse calibration is the manual inventory data provided by the warehouse party.

7. A steel cargo pledge financing Internet of Things supervision system based on a modification of a crown block, characterized in that, The system includes an IoT transformation and data modeling module for the warehouse site and an IoT warehouse supervision module; The Internet of Things transformation and data modeling module of the warehouse site comprises a warehouse area division and measurement sub-module, an overhead traveling crane Internet of Things data acquisition sub-module, an overhead traveling crane behavior judgment model design sub-module, and an in-out warehouse event judgment sub-module The warehouse area division and measurement sub-module is used for virtually dividing the travel range of the overhead traveling crane of the warehouse site and performing area measurement; ensuring that the position of any coordinate in the warehouse can correspond to a virtual area uniquely, and generating a virtual map with coordinates; The overhead traveling crane Internet of Things data acquisition sub-module is used for collecting the hook quality and displacement Internet of Things data of the overhead traveling crane; and combining the quality, coordinate, and time data into a data structure; The overhead traveling crane behavior judgment model design sub-module is used for analyzing the Internet of Things data, restoring the Internet of Things data sequence into overhead traveling crane behavior event data, and counting the corresponding inventory; the Internet of Things data sequence comprises the current X-axis coordinate of the overhead traveling crane, the current Y-axis coordinate of the overhead traveling crane, the current quality data, and the corresponding time data; the behavior event data sequence comprises the behavior event type of loading or unloading, the quality of the goods of the in-out warehouse behavior event, and the occurrence time of the behavior event; The in-out warehouse event judgment sub-module is used for judging whether the Internet of Things data produces a loading or unloading behavior event and the corresponding quality of the goods, obtaining the position and time of the occurrence of the behavior event, and exporting the Internet of Things data sequence into the behavior event data sequence; The in-out warehouse event judgment sub-module is used for judging whether the Internet of Things data produces a loading or unloading behavior event and the corresponding quality of the goods, obtaining the position and time of the occurrence of the behavior event, and exporting the Internet of Things data sequence into the behavior event data sequence; The Internet of Things warehouse supervision module comprises an edge computing sub-module, an Internet of Things data access sub-module, a business data integration sub-module, a warehouse management sub-module, and a risk early warning sub-module; The edge computing sub-module is used for uploading the corresponding Internet of Things data sequence of the overhead traveling crane behavior event to a cloud platform; The Internet of Things data access sub-module is used for realizing the access and subsequent storage of the Internet of Things data, and obtaining the overhead traveling crane in-out warehouse event sequence, wherein each data item represents an in-out warehouse event of goods; The business data integration sub-module is used for obtaining the data related to the electronic warehouse receipt or the goods pledge behavior from the business system of a financial institution; The warehouse management sub-module is used for maintaining the actual inventory of each virtual warehouse area through the Internet of Things data and the in-out warehouse event of goods, with each virtual warehouse area as a unit; The risk early warning sub-module is used for judging whether the warehouse inventory is lower than the weight of the pledged goods, and alarming the relevant person in charge when necessary.

Images