Automobile press line quality monitoring system based on Internet of Things and statistical analysis

Abstract

The invention relates to the technical field of intelligent manufacturing and industrial Internet of Things, and discloses an Internet of Things and statistical analysis-based automobile press line quality monitoring system, which comprises a sensing layer, a transmission layer and an application layer, the sensing layer comprises pressure sensors deployed in a plurality of stamping processes of a stamping line; a data transmission unit is arranged in the transmission layer; the application layer is based on an anomaly monitoring algorithm of statistical design; according to the automobile press line quality monitoring system based on the Internet of Things and statistical analysis, the press line production process is optimized by integrating the Internet of Things technology, the product quality is improved, and the manufacturing industry is promoted to develop towards the intelligent direction. The design is devoted to intelligent upgrading of the automobile stamping production line, the problems of low manual sampling inspection efficiency, missing inspection and the like of an existing stamping production line can be solved, meanwhile, the effect of automatic and continuous storage of PLC processing and real-time output process data can be achieved, and data analysis, modeling and intelligent decision making are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent manufacturing and industrial Internet of Things, and specifically to an automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis. Background Art

[0002] A stamping production line is an automated production system used for sheet metal processing. A large number of metal parts in key components such as automobile bodies, chassis, and engines need to be formed through stamping, such as automobile covers and structural parts. Stamping production lines occupy an extremely important position in automobile manufacturing.

[0003] A stamping production line generally includes multiple stamping processes, and the sheet metal is sequentially processed through multiple stamping processes to achieve stamping and forming. Currently, stamping production lines generally use automated equipment for production, but there are the following problems in product quality monitoring: some production lines rely on manual sampling, which may lead to problems such as missed inspections and low efficiency. Some automated monitoring is mostly limited to PLC processing and real-time output, and process data cannot be automatically and continuously saved, so data analysis, modeling and intelligent decision-making cannot be achieved. Summary of the Invention

[0004] In order to solve the problems raised in the background technology, the present invention is implemented through the following technical solutions: an automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis, including a perception layer, a transmission layer, and an application layer;

[0005] The sensing layer includes pressure sensors deployed in multiple stamping processes of the stamping line, and uses the pressure sensors to collect weight data of the stamping parts output in each process;

[0006] The transmission layer is provided with a data transmission unit, which is used to transmit the weight data to the network server for storage in real time;

[0007] The application layer performs data analysis on the weight of stamping parts based on an abnormality monitoring algorithm designed based on statistics.

[0008] Furthermore, the data transmission unit is a DTU communication module, the pressure sensor is connected to the DTU via an RS-485 interface, and the weight data is sent to the DTU via the Modbus-RTU data protocol.

[0009] Furthermore, the configuration of the DTU is:

[0010] The physical layer uses shielded twisted pair cables to connect the DTU and the 485-A and 485-B differential signal terminals of the pressure sensor. The data link layer protocol uses the Modbus RTU protocol, defining the sensor as a slave and the DTU as a host.

[0011] Furthermore, the DTU uploads the weight data collected by the pressure sensor to the cloud server via a wireless mobile communication network. Based on the upload of the weight data, the configuration of the DTU is as follows:

[0012] The server connection uses TCP transparent transmission mode, the target server address is set to the Alibaba Cloud server's external network address, the port is set to 5000, and the heartbeat interval is set to 60s.

[0013] Furthermore, the statistically designed anomaly monitoring algorithm is specifically:

[0014] By accumulating N batches of normal historical batch data, according to statistical theory, the weight data of stamping parts fluctuates. The weight fluctuations of products produced under the same conditions have regularity. Most of them are concentrated to one value, and at the same time, they are dispersed on both sides of this value. That is, the weight of the finished product of each process is Normal distribution , is the mean, is the standard deviation, For the Process, where:

[0015] ;

[0016] If the weight of the stamping parts in the current process exceeds , indicating that the quality of the stamping part is abnormal.

[0017] Furthermore, the anomaly monitoring steps of the statistical anomaly monitoring algorithm are as follows:

[0018] S1. Calculate the weight statistical distribution of the current process in N historical batches ;

[0019] S2. Collect the weight of the current process ,like exist If the stamping part is within the range, it will be prompted that the stamping part is normal. If it is not within the range, it will be prompted that the stamping part is abnormal.

[0020] S3. Determine the status of S2. If normal, proceed to the next process and include the weight data of this process in the historical data. If abnormal, stop production and check the cause.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] This automotive stamping line quality monitoring system, based on the Internet of Things and statistical analysis, integrates IoT technology to optimize stamping line production processes, improve product quality, and promote the development of intelligent manufacturing. This design is dedicated to the intelligent upgrade of automotive stamping production lines. It can solve the problems of low efficiency and missed inspections in existing manual inspections of stamping lines. At the same time, it can realize the automatic and continuous storage of PLC processing and real-time output process data, realizing data analysis, modeling, and intelligent decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the system architecture of the present invention. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0025] The embodiment of the automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis is as follows:

[0026] See also Figure 1 A quality monitoring system for automobile stamping lines based on the Internet of Things and statistical analysis includes a perception layer, a transmission layer, and an application layer; the perception layer includes pressure sensors deployed in multiple stamping processes of the stamping line, and the pressure sensors are used to collect weight data of the stamping parts output by each process; the transmission layer is provided with a data transmission unit, and the data transmission unit is used to transmit the weight data in real time to a network server for storage; the application layer performs data analysis on the weight of the stamping parts based on a statistically designed abnormality monitoring algorithm.

[0027] It should be noted that the data transmission unit is a DTU communication module, the pressure sensor is connected to the DTU via an RS-485 interface, and the weight data is sent to the DTU via the Modbus-RTU data protocol.

[0028] The configuration of the DTU is as follows: the physical layer uses a shielded twisted pair cable to connect the DTU and the 485-A and 485-B differential signal terminals of the pressure sensor; the data link layer protocol uses the Modbus RTU protocol, defining the sensor as a slave and the DTU as a host.

[0029] It should also be noted that the DTU uploads the weight data collected by the pressure sensor to the cloud server through the wireless mobile communication network. Based on the upload of weight data, the DTU is configured as follows: the server connection adopts TCP transparent transmission mode, the target server address is set to the Alibaba Cloud server external network address, the port is set to 5000, and the heartbeat interval is set to 60s.

[0030] The following is the specific implementation:

[0031] The present invention relies on the Internet of Things technology, in which the application of the Internet of Things is divided into three layers: perception layer, transmission layer, and application layer. These three layers respectively undertake the functions of data acquisition, transmission processing, and application services. In the perception layer, pressure sensors are deployed to collect the weight of the output sheet of each process on the stamping line. The sensor is connected to the DTU through the RS-485 interface, and the sensor weighing data is sent to the DTU through the Modbus-RTU data protocol. The data is then transmitted to the Alibaba Cloud application server in real time through the wireless communication network, and the data is stored in the cloud database to record the weight of the stamping parts after different processes. According to the anomaly monitoring algorithm, the application server analyzes the weight before and after each process and identifies anomalies. For a stamping production line with 4 processes (OP10-OP40), the overall architecture of the system is as follows Figure 1 shown.

[0032] The perception layer is the bottom layer of the IoT, providing the physical foundation, interacting directly with the environment, and collecting physical information in real time. This system deploys a high-precision pressure sensor system in four stamping processes, OP10, OP40, and OP50. It uses a JHBM-H1 pressure sensor (range 0-100 kg, nonlinearity ±0.03% FS, IP66 protection rating). The pressure sensor converts the pressure signal into an analog electrical signal, which is then converted to a digital signal via an RS-485 interface via a digital transmitter. The pressure sensor measures the output weight of processes 1, 2, 3, and 4, respectively. Data aggregation of four sensor networks is achieved through the Modbus-RTU protocol. The baud rate of the sensor is set to 9600bps, the data frame format is: 1 start bit, 8 data bits, 1 stop bit, no parity, and the sensor address assignment is: OP10 output: 0x01, OP20 output: 0x02, OP30 output: 0x03, OP40 output: 0x04. The DTU acts as the master station and sends query commands to the sensor system to collect weight data.

[0033] The transport layer provides information transmission channels for IoT applications, integrating wide-area transmission technologies such as wired and wireless communication networks to upload perception layer data to the cloud. This system uses the USER-G786 industrial-grade 4G DTU from the UIoT as its core communication module. Its main features include a built-in dual-mode, fully Netcom 4G Cat1 communication module and support for a wide operating temperature range of -35°C to 75°C, meeting the requirements of industrial scenarios for device stability and adaptability to complex environments. It also provides a standard RS-485 interface for direct connection to industrial fieldbus devices, supports a wide voltage range of 9-36V DC power supply, and maintains a standby power consumption of ≤1.2W, meeting energy efficiency standards for industrial IoT devices.

[0034] To ensure reliable data exchange between the DTU and the pressure sensor, the following configuration is required: At the physical layer, use a shielded twisted-pair cable to connect the DTU to the sensor's 485-A and 485-B differential signal terminals. At the data link layer, use Modbus RTU, define the sensor as a slave and the DTU as a master. Set the data frame format to 1 start bit, 8 data bits, 1 stop bit, no parity check, and a baud rate of 9600 bps. To enable the DTU to upload sensor data to the cloud via a wireless mobile communication network, the following configuration is required: Use TCP transparent transmission mode for the server connection, set the target server address to the Alibaba Cloud server's external network address, set the port to 5000, and set the heartbeat interval to 60 seconds.

[0035] The application layer is the top layer of the IoT architecture. Data is aggregated from the network transmission layer and processed, analyzed, and stored at this layer. This system design uses process control technology to analyze the weight of stamped parts across multiple processes. Based on the normal distribution characteristics of historical data, an algorithm was designed to detect anomalies. Simultaneously, a server-side program was developed to store, monitor, and analyze data, providing support for industrial control and production monitoring, ensuring stable, high-quality operation across all processes of the automotive stamping line.

[0036] It should be noted that the statistically designed abnormality monitoring algorithm is specifically as follows: by accumulating N batches of normal historical batch data (N≥40), according to statistical theory, the weight data of stamping parts fluctuates, and the weight fluctuations of products produced under the same conditions have regularity, most of which are concentrated to one value, and at the same time dispersed on both sides of this value, that is, the weight of the finished product of each process is Normal distribution , is the mean, is the standard deviation, For the Process, where:

[0037] ;

[0038] The 3σ principle is often used in process control and quality management. Most of the observed values will fall within the 3σ range. The probability of data points outside this range is very small. If the weight value of the stamping part in the current process exceeds , indicating that the quality of the stamping parts is abnormal. Through this statistical analysis, an early warning can be given that the stamping parts in the current process have quality defects.

[0039] The anomaly detection steps of the statistical anomaly detection algorithm are:

[0040] Normal data of N historical batches has been accumulated. In a certain production, each process OP(i) (i=1, 2, 3, 4) is analyzed in turn:

[0041] S1. Calculate the weight statistical distribution of the current process in N historical batches ;

[0042] S2. Collect the weight of the current process ,like exist If the stamping part is within the range, it will be prompted that the stamping part is normal. If it is not within the range, it will be prompted that the stamping part is abnormal.

[0043] S3. Determine the status of S2. If normal, proceed to the next process and include the weight data of this process in the historical data. If abnormal, stop production and check the cause.

[0044] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. An automobile press line quality monitoring system based on the Internet of Things and statistical analysis, characterized by: Includes perception layer, transport layer, and application layer; The sensing layer includes pressure sensors deployed in multiple stamping processes of the stamping line, and uses the pressure sensors to collect weight data of the stamping parts output in each process; The transmission layer is provided with a data transmission unit, which is used to transmit the weight data to the network server for storage in real time; The application layer performs data analysis on the weight of stamping parts based on an abnormality monitoring algorithm designed based on statistics.

2. The automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis according to claim 1 is characterized in that: The data transmission unit is a DTU communication module. The pressure sensor is connected to the DTU via an RS-485 interface and sends weight data to the DTU via the Modbus-RTU data protocol.

3. The automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis according to claim 1 is characterized in that: The configuration of the DTU is: The physical layer uses shielded twisted pair cables to connect the DTU and the 485-A and 485-B differential signal terminals of the pressure sensor. The data link layer protocol uses the Modbus RTU protocol, defining the sensor as a slave and the DTU as a host.

4. The automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis according to claim 3 is characterized by: The DTU uploads the weight data collected by the pressure sensor to the cloud server via the wireless mobile communication network. Based on the upload of weight data, the configuration of the DTU is as follows: The server connection uses TCP transparent transmission mode, the target server address is set to the Alibaba Cloud server's external network address, the port is set to 5000, and the heartbeat interval is set to 60s.

5. The automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis according to claim 1 is characterized in that: The statistically designed anomaly monitoring algorithm is specifically: By accumulating N batches of normal historical batch data, according to statistical theory, the weight data of stamping parts fluctuates. The weight fluctuations of products produced under the same conditions have regularity. Most of them are concentrated to one value, and at the same time, they are dispersed on both sides of this value. That is, the weight of the finished product of each process is Normal distribution , is the mean, is the standard deviation, For the Process, where: ； If the weight of the stamping parts in the current process exceeds , indicating that the quality of the stamping part is abnormal.

6. The automobile stamping line quality monitoring system based on the Internet of Things and statistical analysis according to claim 5 is characterized in that: The anomaly detection steps of the statistical anomaly detection algorithm are: S1. Calculate the weight statistical distribution of the current process in N historical batches ; S2. Collect the weight of the current process ,like exist If the stamping part is within the range, it will be prompted that the stamping part is normal. If it is not within the range, it will be prompted that the stamping part is abnormal. S3. Determine the status of S2. If normal, proceed to the next process and include the weight data of this process in the historical data. If abnormal, stop production and check the cause.

Images