Intelligent early warning method based on automatic stereoscopic warehouse

Abstract

An intelligent early warning method based on an automatic three-dimensional library comprises the steps that firstly, all data in a management system of the automatic three-dimensional library are collected, and then an action library is designed according to equipment needing to be configured for forming normal operation of the automatic three-dimensional library and the collected data; carrying out picture collection on the normally operating automatic three-dimensional library equipment, and then carrying out preprocessing to construct a picture environment library; the training learning model selects a proper operation action from an action library, and constructs and trains an early warning training learning model based on a PID algorithm, according to the characteristics of the automatic stereo library, based on an operation control algorithm and a constructed picture environment library and in combination with an early warning decision strategy of the automatic stereo library; and finally, the early warning training learning model is connected to a management system of the automatic stereo library, equipment operation pictures shot in the automatic stereo library are received in real time and processed, and safety early warning is timely sent out, so that safety production of the automatic stereo library is guaranteed. Meanwhile, possible risks are checked, and the early warning effect is further improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent decision-making technology, and in particular to an intelligent early warning method based on an automated three-dimensional warehouse. Background Art

[0002] With the development of artificial intelligence and integrated sensor technology, automated stereoscopic warehouses for single-roll storage of materials have also emerged. The management system controls the conveying system, packaging system, and code scanner. The conveying system includes a coil conveyor, a pallet conveyor, a transport vehicle, a storage conveyor, and a stacker on the same transmission line, and also includes an outbound conveyor on another transmission line with the stacker; the packaging system includes a packaging line set at the front end of the coil conveyor and a coding system used in conjunction with the packaging line; the code scanner is used in conjunction with the conveying system; the stereoscopic warehouse body is the carrier of the conveyed materials, and the normal operation of the automated stereoscopic warehouse involves various links related to the automated stereoscopic warehouse and various steps of the operator's operation. Therefore, it is extremely important to study the operating status of the automated stereoscopic warehouse and each system, and monitor, diagnose and isolate faults for them, which is extremely important for ensuring the safe production of the automated stereoscopic warehouse. Summary of the invention

[0003] The technical problem solved by the present invention is to provide an intelligent early warning method based on an automated three-dimensional warehouse to solve the problems in the above-mentioned background technology.

[0004] The technical problem solved by the present invention is achieved by adopting the following technical solutions: An intelligent early warning method based on an automated stereoscopic warehouse, the specific steps are as follows: 1) Data collection The data acquisition module uses bus technology to collect all data in the management system of the automated three-dimensional warehouse. In order to ensure the safety and reliability of data acquisition during the data acquisition process, the data acquisition module adopts a one-way flow mode, which will not interfere with the management system of the automated three-dimensional warehouse; 2) Real-time storage The data acquisition module stores the collected data in real time; 3) Design action library The action library is designed according to the equipment and data collected to form the normal operation of the automated three-dimensional warehouse. The equipment that forms the normal operation of the automated three-dimensional warehouse includes a management system, a conveying system, a packaging system, and a barcode scanner. The conveying system includes a coil conveyor, a pallet conveyor, a transport vehicle, an inbound conveyor, a stacker, and an outbound conveyor on another transmission line with the stacker; the packaging system includes a packaging line arranged at the front end of the coil conveyor and a coding system used in conjunction with the packaging line; The equipment that constitutes the normal operation of the automated three-dimensional warehouse is split according to the operating status of each individual equipment and numbered in sequence to generate an action library. The action library includes basic operating actions: ① uniform speed straight line operation; ② interval straight line operation; ③ minimum deceleration straight line operation; ④ maximum acceleration straight line operation; ⑤ maximum climbing operation; ⑥ maximum diving operation; 4) Image collection Capture images of the normally operating automated stereoscopic storage equipment, and then pre-process the captured images to build an image environment library; pre-processing includes normalizing the pixels and size of the images; 5) Build an early warning training learning model The training learning model selects the appropriate operation action from the action library in step 3), and based on the PID algorithm, converts the conveying control amount obtained from the automated stereoscopic warehouse conveying system into the conveying position. Then, according to the characteristics of the automated stereoscopic warehouse, based on the operation control algorithm and the constructed image environment library, combined with the automated stereoscopic warehouse early warning decision strategy, the early warning training learning model is constructed by learning and training the automated stereoscopic warehouse early warning decision; The early warning training and learning model includes the design of neural network structure and the selection of training and learning algorithms; The training learning model adopts the PPO deep reinforcement learning method, and the neural network structure design includes the automatic three-dimensional warehouse early warning decision network structure design and evaluation automatic three-dimensional warehouse early warning decision network structure design; 6) Training warning training learning model Input various equipment operation pictures into the early warning training learning model to train the early warning training learning model, and visualize the training results until the early warning training learning model realizes all requirements for automatic three-dimensional library early warning decision-making; 7) Online warning The early warning training learning model is connected to the management system of the automated stereoscopic warehouse. The early warning training learning model receives real-time equipment operation pictures taken in the automated stereoscopic warehouse, and then processes the pictures. When dangerous or abnormal information is found in the pictures, a safety warning is issued in time through the management system.

[0005] In the present invention, in step 5), the upper and lower operating limits of the operation of the automated stereoscopic warehouse equipment are set in the automated stereoscopic warehouse early warning decision.

[0006] In the present invention, in step 5), the automated stereoscopic warehouse operation action and the image environment inventory in the automated stereoscopic warehouse early warning decision have a one-to-one correspondence relationship, which is used to determine the automated stereoscopic warehouse operation status.

[0007] In the present invention, in step 7), the equipment operation picture of the automated stereoscopic warehouse is taken by streaming media or camera.

[0008] In the present invention, in step 7), the historical data of the automated stereoscopic warehouse equipment is pre-stored in the early warning training learning model, and the historical data of the automated stereoscopic warehouse equipment is regularly compared with the existing equipment data to check for possible risks and further improve the early warning effect.

[0009] Beneficial effects: The present invention connects the constructed early warning training learning model to the management system of the automated warehouse, which can monitor the operating status of the automated warehouse and its various systems, diagnose faults and isolate faults. When dangerous or abnormal information is found in the operating picture, a safety warning is issued in time through the management system, thereby ensuring the safe production of the automated warehouse; at the same time, possible risks are checked to further improve the early warning effect. DETAILED DESCRIPTION

[0010] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further described below in conjunction with specific embodiments.

[0014] An intelligent early warning method based on an automated stereoscopic warehouse, the specific steps are as follows: 1) Data collection The data acquisition module uses bus technology to collect all data in the management system of the automated three-dimensional warehouse. In order to ensure the safety and reliability of data acquisition during the data acquisition process, the data acquisition module adopts a one-way flow mode, which will not interfere with the management system of the automated three-dimensional warehouse; 2) Real-time storage The data acquisition module stores the collected data in real time; 3) Design action library The action library is designed according to the equipment and data collected to form the normal operation of the automated three-dimensional warehouse. The equipment that forms the normal operation of the automated three-dimensional warehouse includes a management system, a conveying system, a packaging system, and a barcode scanner. The conveying system includes a coil conveyor, a pallet conveyor, a transport vehicle, an inbound conveyor, a stacker, and an outbound conveyor on another transmission line with the stacker; the packaging system includes a packaging line arranged at the front end of the coil conveyor and a coding system used in conjunction with the packaging line; The equipment that constitutes the normal operation of the automated three-dimensional warehouse is split according to the operating status of each individual equipment and numbered in sequence to generate an action library. The action library includes basic operating actions: ① uniform speed straight line operation; ② interval straight line operation; ③ minimum deceleration straight line operation; ④ maximum acceleration straight line operation; ⑤ maximum climbing operation; ⑥ maximum diving operation; 4) Image collection Capture images of the normally operating automated stereoscopic storage equipment, and then pre-process the captured images to build an image environment library; pre-processing includes normalizing the pixels and size of the images; 5) Build an early warning training learning model The training learning model selects the appropriate operation action from the action library of step 3), based on the PID algorithm, the linear speed control is controlled based on the difference between the tangential overload instruction and the actual tangential overload of the conveying system, and the pitch control is controlled based on the difference between the normal overload instruction and the actual normal overload of the conveying system, and the conveying control amount obtained from the automated stereoscopic warehouse conveying system is converted into the conveying position; then according to the characteristics of the automated stereoscopic warehouse, based on the operation control algorithm and the constructed image environment library, combined with the automated stereoscopic warehouse early warning decision strategy, through the training learning model learning, training the automated stereoscopic warehouse early warning decision, to build an early warning training learning model; The early warning training and learning model includes the design of neural network structure and the selection of training and learning algorithms; The training learning model adopts the PPO deep reinforcement learning method, and the neural network structure design includes the automatic three-dimensional warehouse early warning decision network structure design and evaluation automatic three-dimensional warehouse early warning decision network structure design; 6) Training warning training learning model Input various equipment operation pictures into the early warning training learning model to train the early warning training learning model, and visualize the training results until the early warning training learning model realizes all requirements for automatic three-dimensional library early warning decision-making; 7) Online warning The early warning training learning model is connected to the management system of the automated stereoscopic warehouse. The early warning training learning model receives real-time equipment operation pictures taken in the automated stereoscopic warehouse, and then processes the pictures. When dangerous or abnormal information is found in the pictures, a safety warning is issued in time through the management system.

[0015] In this embodiment, in step 5), the upper and lower operating limits of the operation of the automated stereoscopic warehouse equipment are set in the automated stereoscopic warehouse early warning decision.

[0016] In this embodiment, in step 5), there is a one-to-one correspondence between the automated stereoscopic warehouse operation action and the image environment inventory in the automated stereoscopic warehouse early warning decision, which is used to determine the automated stereoscopic warehouse operation status.

[0017] In this embodiment, in step 7), the equipment operation pictures of the automated stereoscopic warehouse are taken by streaming media or a camera.

[0018] In this embodiment, in step 7), the historical data of the automated stereoscopic warehouse equipment is pre-stored in the early warning training learning model, and the historical data of the automated stereoscopic warehouse equipment is regularly compared with the existing equipment data to check for possible risks and further improve the early warning effect.

Claims

1. An intelligent early warning method based on an automated stereoscopic warehouse, characterized in that: The specific steps are as follows: 1) Data collection The data acquisition module collects all data in the management system of the automated stereoscopic warehouse; 2) Real-time storage The data acquisition module stores the collected data in real time; 3) Design action library According to the equipment and data collected for the normal operation of the automated three-dimensional warehouse, the equipment that constitutes the normal operation of the automated three-dimensional warehouse is divided according to the operating status of each device and numbered in sequence to generate an action library. The action library includes basic operating actions: ① uniform speed straight line operation; ② interval straight line operation; ③ minimum deceleration straight line operation; ④ maximum acceleration straight line operation; ⑤ maximum climbing operation; ⑥ maximum diving operation; 4) Image collection Collect pictures of the normally operating automated stereoscopic storage equipment, and then pre-process the collected pictures to build a picture environment library; 5) Build an early warning training learning model The training learning model selects the appropriate operation action from the action library in step 3), and based on the PID algorithm, converts the conveying control amount obtained from the automated stereoscopic warehouse conveying system into the conveying position. Then, according to the characteristics of the automated stereoscopic warehouse, based on the operation control algorithm and the constructed image environment library, combined with the automated stereoscopic warehouse early warning decision strategy, the early warning training learning model is constructed by learning and training the automated stereoscopic warehouse early warning decision; 6) Training warning training learning model Input various equipment operation pictures into the early warning training learning model to train the early warning training learning model, and visualize the training results until the early warning training learning model realizes all requirements for automatic three-dimensional library early warning decision-making; 7) Online warning The early warning training learning model is connected to the management system of the automated stereoscopic warehouse. The early warning training learning model receives real-time equipment operation pictures taken in the automated stereoscopic warehouse, and then processes the pictures. When dangerous or abnormal information is found in the pictures, a safety warning is issued in time through the management system.

2. According to claim 1, an intelligent early warning method based on an automated stereoscopic warehouse is characterized in that: In step 1), the data acquisition module adopts a unidirectional flow mode.

3. The intelligent early warning method based on the automated stereoscopic warehouse according to claim 1 is characterized in that: In step 4), the preprocessing includes normalizing the pixels and size of the image.

4. The intelligent early warning method based on the automated stereoscopic warehouse according to claim 1 is characterized in that: In step 5), the early warning training learning model includes the neural network structure design and the selection of the training learning algorithm. The neural network structure design includes the automated three-dimensional warehouse early warning decision network structure design and the evaluation of the automated three-dimensional warehouse early warning decision network structure design.

5. The intelligent early warning method based on the automated stereoscopic warehouse according to claim 1 is characterized in that: In step 5), the upper and lower limits of the operation of the automated stereoscopic warehouse equipment are set in the automated stereoscopic warehouse early warning decision, and the automated stereoscopic warehouse operation actions have a one-to-one correspondence with the image environment inventory.

6. The intelligent early warning method based on the automated stereoscopic warehouse according to claim 1 is characterized in that: In step 7), the equipment operation picture of the automated stereoscopic warehouse is captured by streaming media or camera.

7. The intelligent early warning method based on the automated stereoscopic warehouse according to claim 1 is characterized in that: In step 7), the historical data of the automated stereoscopic warehouse equipment is pre-stored in the early warning training learning model, and the historical data of the automated stereoscopic warehouse equipment is compared with the existing equipment data on a regular basis.