Intelligent motion recognition and dynamic load weighing method for crown block

Abstract

The application discloses a kind of intelligent action recognition and dynamic object weight measurement method for crown block, comprising the following steps: S1, the real-time data of bottom layer hardware in steelmaking site is collected with frequency f1;S2, a vector W with window size n is taken out from the real-time data list of the weight of the crown block with frequency f2, the variance Wvar of the vector is calculated, the weight is locked and the locking time is recorded, and the size of Wvar and the preset value Vset1 is compared, if Wvar>Vset1, the action starts, otherwise the action does not start;S3, if the action does not start, repeat S2 step until the action starts;If the action starts, record xt, yt as the coordinates xb, yb of the start of action, and obtain the locked object weight Wlockb as the object weight at the start of action, record tb as the time at the start of action;S4, if the action starts, update W and Wvar, judge whether to exit the action;Satisfy any one of conditions 2 to 6, and satisfy condition 1, then the action ends.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of overhead crane logistics, and particularly relates to an intelligent action recognition and dynamic load measurement method for an overhead crane. BACKGROUND

[0002] In overhead crane logistics, the carrier involved is an overhead crane, and the main hoisted objects involved are steel ladles, scrap steel hoppers, iron ladles and steel coils. The basic requirement of overhead crane logistics is to track and manage the circulation of the hoisted objects of the overhead crane. In order to meet this requirement, it is necessary to know when and where the hoisted objects of the overhead crane are hoisted, what action is performed on the hoisted objects, and obtain the weight of the hoisted objects. The overhead crane hoisted object action recognition and hoisted object load measurement method is to meet the basic requirement of overhead crane logistics. Therefore, the present application provides an intelligent action recognition and dynamic load measurement method for an overhead crane to solve the above problems. SUMMARY

[0003] The present application aims to provide an intelligent action recognition and dynamic load measurement method for an overhead crane to solve the problems raised in the background.

[0004] To achieve the above-mentioned purpose, the present application provides the following technical solution: an intelligent action recognition and dynamic load measurement method for an overhead crane, comprising the following steps:

[0005] S1, collecting real-time data of the bottom layer hardware of the steelmaking site at a frequency f1;

[0006] S2, taking a vector W with a time window size of n from the real-time data list of the load of the overhead crane at a frequency f2, calculating the variance Wvar of the vector, locking the load and recording the locking time, and comparing the size of Wvar and a preset value Vset1. If Wvar>Vset1, the action starts, otherwise the action does not start;

[0007] S3, if the action does not start, repeating the S2 step until the action starts; if the action starts, recording xt, yt as the coordinates xb, yb of the start of the action, and obtaining the locked load Wlockb as the load at the start of the action, and recording tb as the time at the start of the action;

[0008] S4, if the action starts, updating W and Wvar, and determining whether to exit the action; if any one of conditions 2 to 6 is met and condition 1 is met, the action ends, otherwise the action does not end and the S4 step is repeated until the action ends. If the action ends, recording xt, yt as the coordinates xa, ya of the end of the action, and obtaining the locked load Wlocka as the load at the end of the action, and recording ta as the time at the end of the action;

[0009] S5, after the action, through Wlockb and Wlocka to determine the action type, then the next detection whether there is a start of action, that is, repeat S2 step.

[0010] Preferably, the real-time data in S1 mainly includes the real-time coordinates xi axis, yi axis of a steelmaking crane and the weight wi of the crane.

[0011] Preferably, the calculation formula of W and Wvar in S2 is W = w t-n ......w t-1 , w t ,

[0012] Preferably, the process of locking the weight in S2 includes the following steps:

[0013] Q1, update the real-time data xi, yi, W of the crane;

[0014] Q2, determine whether the number of W is greater than n, when the number of W is greater than or equal to n, update Wvar and execute Q3, when the number of W is less than n, continue to execute Q1;

[0015] Q3, determine the size relationship between Wvar and Vset, use multi-level variance Wlock = (W-min(W)-max(W)) / (n-2) to record the locking time tlock;

[0016] Q4, multi-level weight locking is completed, enter the next round of real-time data.

[0017] Preferably, the locking time tlock is tlock1, tlock2, tlock3..., the Wlock is Wlock1, Wlock2, Wlock3..., the Vset is Vset1, Vset2, Vset3..., and the judgment condition of the size relationship between Wvar and Vset is whether Wvar <= Vset.

[0018] Preferably, the condition in S4 is as follows:

[0019] 1. The weight of the crane is relatively stable, that is, Wvar < Vset2;

[0020] 2. The x-axis coordinate of the crane moves more than the preset value Xset relative to the x-axis coordinate at the start of the action, that is, abs(xa-xt) >= Xset;

[0021] 3. The y-axis coordinate of the crane moves more than the preset value Yset relative to the y-axis coordinate at the start of the action, that is, abs(ya-yt) >= Yset;

[0022] 4, the current time tt compared with the action start time tb exceeds the preset value Tset, that is, tt-tb >= Tset;

[0023] 5, the fast lifting and lowering, the average value of the weight of the object is about 0, that is,

[0024] 6, the fast lifting and lowering, the average value of the weight of the object is about the maximum value of the re-packaging, that is,

[0025] Compared with the prior art, the beneficial effects of the present application are:

[0026] 1, the weight of the object measured by the present application is more accurate, and the overall more stable measured weight of the object can be obtained within a period of time, thereby avoiding obtaining a locally stable weight of the object;

[0027] 2, the present application can avoid the problem of losing action caused by the on-site half-pack lifting;

[0028] 3, the present application judges the stability of the weight of the object by the variance, and can accurately identify the start of each action, and ensure the integrity of a series of actions;

[0029] 4, the present application optimizes the lifting and lowering speed, and can quickly exit the action for a specific action, thereby avoiding the action cross of multiple cranes caused by the delay of ending the action. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is the overall flowchart of action recognition in the present application;

[0031] Figure 2 is the weight locking flowchart in the present application;

[0032] Figure 3 is the weight obtaining flowchart when the action starts in the present application;

[0033] Figure 4 is the weight obtaining flowchart when the action ends in the present application;

[0034] Figure 5 is the action type judgment flowchart in the present application. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0036] EMBODIMENT

[0037] Referring to Figures 1-5 The embodiment provides an intelligent action recognition and dynamic weight measurement method for a crown block, taking a crown block as an example, and comprising the following steps:

[0038] S1, collecting real-time data of a bottom-layer hardware in a steelmaking site at a frequency f1, wherein the real-time data mainly comprises real-time coordinates xi-axis, yi-axis of a certain steelmaking crown block in the site and a weight wi of a load hoisted by the crown block;

[0039] S2, taking a vector W with a time window size n from a real-time data list of the weight of the load hoisted by the crown block at a frequency f2, calculating a variance Wvar of the vector, locking the weight and recording a locking time, such as Figure 2 The multi-stage variance locks the weight, and the specific process comprises the following steps:

[0040] Q1, updating the real-time data xi, yi and W of the crown block;

[0041] Q2, judging whether the number of W is greater than n, when the number of W is greater than or equal to n, updating Wvar and executing Q3, when the number of W is less than n, continuing to execute Q1;

[0042] Q3, judging the size relationship between Wvar and Vset, and recording the locking time tlock by using the multi-stage variance Wlock=(W-min(W)-max(W)) / (n-2);

[0043] Q4, ending the multi-stage weight locking, and entering the next round of real-time data.

[0044] The locking time tlock in the above steps are tlock1, tlock2, tlock3..., Wlock are Wlock1, Wlock2, Wlock3..., Vset are Vset1, Vset2, Vset3..., and the judgment condition of the size relationship between Wvar and Vset is whether Wvar<=Vset, wherein Vset1

[0045] Then, the size of Wvar and a preset value Vset1 is compared, if Wvar>Vset1, the action starts, otherwise the action does not start. The current time is t, and the calculation formulas of W and Wvar are W=w t-n ......w t-1 , w t ,

[0046] S3, if the action has not started, repeat S2 until the action starts; if the action has started, record xt, yt as the coordinates xb, yb at which the action started, and obtain the locked weight Wlockb at which the action started, record tb as the time at which the action started; as shown in Figure 3 as long as the time at which the weight is locked is between the current action start time tb and the last action end time ta_last, it is a usable locked weight, and the most stable one is obtained by covering the less stable weight with the more stable weight;

[0047] S4, if the action has started, update W and Wvar, and determine whether to exit the action; if any one of conditions 2 to 6 is met, and condition 1 is met, the action ends, otherwise the action does not end and repeats S4 until the action ends. If the action ends, record xt, yt as the coordinates xa, ya at which the action ends, and obtain the locked weight Wlocka at which the action ends, record ta as the time at which the action ends; as shown in Figure 4 as long as the time at which the weight is locked is between the current action end time ta and the last action start time tb_last, it is a usable locked weight, and the most stable one is obtained by covering the less stable weight with the more stable weight.

[0048] The conditions in step S4 are as follows:

[0049] 1. The crane weight is relatively stable, i.e. Wvar < Vset2;

[0050] 2. The x-axis coordinate of the crane moves more than a preset value Xset relative to the x-axis coordinate at which the action starts, i.e. abs(xa-xt) >= Xset;

[0051] 3. The y-axis coordinate of the crane moves more than a preset value Yset relative to the y-axis coordinate at which the action starts, i.e. abs(ya-yt) >= Yset;

[0052] 4. The current time tt exceeds a preset value Tset compared to the action start time tb, i.e. tt-tb >= Tset;

[0053] 5. The average weight during fast lowering is about 0, i.e.

[0054] 6. The average weight during fast re-wrapping is about the maximum re-wrapping weight, i.e.

[0055] S5, after the action ends, determine the action type by Wlockb and Wlocka, as shown in Figure 5As shown, the judgment of the action type first judges whether it is the exchange object or the exchanged object, then judges whether it is the lifting or the putting, and then judges whether it is the empty container or the heavy container; for the action of the exchange object or the exchanged object, Yset needs to be expanded, and it is returned to the S4 step to continue to judge whether the action is exited; for the non-exchange object or the exchanged object action, Yset is reset, and then the next detection of whether the action is started is performed, that is, the S2 step is repeated.

[0056] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent motion recognition and dynamic load weighing for a crown block, characterized in that, It comprises the following steps: S1, collecting real-time data of the bottom hardware of the steelmaking site at a frequency f1; S2, taking a vector W with a time window size of n from the real-time data list of the crane load, calculating the variance Wvar of the vector, locking the load and recording the locking time, and comparing the size of Wvar and the preset value Vset1, if Wvar>Vset1, the action starts, otherwise the action does not start; S3, if the action does not start, repeat the step S2 until the action starts; if the action starts, record xt, yt as the coordinates xb, yb of the action start, and obtain the locked load Wlockb as the load at the action start, and record tb as the time at the action start; S4, if the action starts, update W and Wvar, and judge whether to exit the action; if any one of conditions 2 to 6 is met and condition 1 is met, the action ends, and the specific conditions are as follows: 1, the crane load is relatively stable, i.e. Wvar<Vset2; 2, the x-axis coordinate of the crane moves more than the preset value Xset relative to the x-axis coordinate at the start of the action, i.e. abs(xa-xt)>=Xset; 3, the y-axis coordinate of the crane moves more than the preset value Yset relative to the y-axis coordinate at the start of the action, i.e. abs(ya-yt)>=Yset; 4, the current time tt exceeds the preset value Tset compared with the action start time tb, i.e. tt-tb>=Tset; 5. The average weight of the fast release and suspension is about 0, i.e. ; 6, fast re-pack hoisting, the average weight of the object is about the maximum value of the re-pack, that is ; otherwise the action does not end and the step S4 is repeated until the action ends, if the action ends, record xt, yt as the coordinates xa, ya at the end of the action, and obtain the locked load Wlocka as the load at the end of the action, and record ta as the time at the end of the action; S5, after the action ends, the action type is judged by Wlockb and Wlocka, the action type is first judged whether it is a material or a material to be exchanged, then whether it is lifting or lowering, and then whether it is an empty container or a heavy container; for the action of exchanging material or material to be exchanged, Yset needs to be expanded, and the step S4 is returned to continue to judge whether to exit the action; for the non-exchanging material or material to be exchanged action, Yset is reset, and then the next detection whether the action starts is performed, i.e. the step S2 is repeated.

2. The method for intelligent motion recognition and dynamic load measurement of a crown block according to claim 1, characterized in that: The real-time data in the S1 mainly includes the real-time coordinates xi, yi of a steelmaking crane on site and the load wi of the crane.

3. The method of claim 1, wherein: The calculation formula of W and Wvar in S2 is respectively .

4. The method of claim 3, wherein: The process of locking the load in the S2 comprises the following steps: Q1, updating the real-time data xi, yi, W of the crane; Q2, judging whether the number of W is greater than n, when the number of W is greater than or equal to n, updating Wvar and executing Q3, when the number of W is less than n, continuing to execute Q1; Q3, judging the size relationship between Wvar and Vset, and recording the locking time tlock by using the multi-level variance Wlock=(W-min(W)-max(W)) / (n-2); Q4, the multi-level load locking ends, and the next round of real-time data is entered.

5. The method of claim 4, wherein: The lock time tlock is tlock1, tlock2, tlock3..., respectively, the Wlock is Wlock1, Wlock2, Wlock3..., respectively, the Vset is Vset1, Vset2, Vset3..., respectively, and the judgment condition of the size relationship between the Wvar and the Vset is whether Wvar<=Vset.

Images