With the continuous promotion of "industry 4.0" and "made in China 2025", the demand of enterprises for flexible manufacturing is also increasing. Customized production of products with more varieties and less quantity has become a significant feature of intelligent manufacturing of modern production enterprises. The transportation of materials and goods required in the production process of enterprises has become an important part of automatic production and manufacturing. How to achieve high efficiency, automation and intelligence, and realize the automation and intellectualization of a series of work such as production, manufacturing and transportation is the goal that people need to explore and study all the time. Automatic guided vehicle (AGV) is a kind of transport vehicle without driver's field operation in logistics transportation automation because of its autonomous positioning and navigation device, which can drive according to the planned route, has the function of safety protection and load transfer It has been widely used in factory automation product production and intelligent manufacturing. It has become a necessary means of automatic handling in product production, intelligent manufacturing and logistics system, which greatly improves the efficiency of workpiece transportation between production and processing processes and reduces the production cost.
1 control system scheme
The control system scheme of this study is mainly suitable for AGV, which is used in the production, manufacturing and assembly of products in the production process. According to the order of raw materials and parts processing required by product production, the automatic guided robot AGV selects goods from the raw material warehouse, and automatically guides and automatically transports to the designated position. In order to complete these actions, AGV must be able to reliably transport raw materials, semi-finished products or parts to the designated station according to the handling instructions input by the production dispatcher on the server side, and can continuously transport multiple stations according to the instructions. In addition, during transportation, AGV should have safety protection measures, such as emergency stop, obstacle avoidance and safety touch in case of emergency, so as to ensure the safety of personnel and equipment on the production site, meet the automation and flexibility of enterprise production process, and lay a good foundation for intelligent manufacturing.
2. Function and composition of control system
All kinds of subtle actions of the automatic guidance robot are completed by the CPU central processing unit of the robot's brain, and the program body is controlled by each component, and the programmed control is realized by computer technology and program . The main functions of AGV control system of AGV include:
(1) The wireless communication between the server and AGV is realized by wireless communication module. The production dispatcher inputs the handling instructions through the HMI of the upper computer on the server side, and transmits the instructions to the AGV through the wireless communication module The central processing unit (CPU) processes the input instructions and controls the AGV to start running and make the specified actions.
(2) The AGV starts to run and plans the path automatically, and collects the path information in real time and uploads it back to the server. The AGV status information is displayed on the HMI of the upper computer, and the steering deviation and cumulative error are corrected.
(3) In the process of material handling, the position specified by the control command can be reliably and accurately positioned to realize the material moving in and out. After the specified action of the station is completed, according to the instruction input by the production dispatcher in (1), continue to go to the next station or return to the origin of AGV to wait for the operation of the next task.
(4) The battery voltage of AGV power supply of AGV is monitored in real time. When the voltage is lower than the specified value, an alarm will be sent out to prompt the battery to be charged.
2.1 control system composition
The control system is divided into three levels: user layer, execution layer and perception layer, which are connected by network. The block diagram of AGV control system of AGV is shown in Fig. 1, which describes the detailed connection among the three levels of the control system and the control function units. The user layer refers to the human-computer interaction unit, the executive layer includes the central processing unit, navigation unit, driving unit, power supply unit and security unit, and the network layer refers to the communication network between the communication unit and each functional unit.
2.2 control unit functions of control system
Central processing unit: the AGV control system of AGV adopts PLC as the main controller, which is the core control unit of the whole control system. Through the analysis and processing of data information collected by sensors, the corresponding judgment is made , and the logic control and motion control of AGV are completed.
The control program of AGV control system of AGV is stored in the main controller PLC. The production dispatchers input work instructions from the man-machine interface of the upper computer. After the instructions are analyzed, the main controller PLC controls the relevant control units and completes the tasks specified in the instructions. Human computer interaction unit: there are two main human-computer interaction interfaces for displaying AGV of AGV. One is the human-computer interaction interface on the server side, and the other is the human-computer interaction interface on the AGV body of the automatic guidance robot. The functions of the two interfaces are the same, which are developed by Kingview software, including status screen, operation picture, alarm picture, operation record, etc.
Navigation unit: collect and process the environmental data, determine the current position and attitude of AGV, and send the calculated deviation from the target position and attitude back to the main controller PLC of AGV, which is used to control the system to give instructions to the drive system and adjust the position and attitude deviation.
Driving unit: walking motor and steering motor are used to realize the walking and steering of AGV. The low-voltage DC motor with band brake function is selected as the motor, and the low-voltage DC servo driver is used to ensure the navigation and positioning accuracy. As shown in Fig. 2, the front wheel is used as the steering driving wheel, which is driven by the steering motor, and the rear two wheels are the walking driving wheels driven by the walking motor. According to the path planning signal sent by the main controller PLC, the motor drive module drives the motor to complete the forward, backward and steering actions.
In Figure 2, l is the length between the driving wheels of AGV, i.e. the track width; P is the center point of the track; R is the radius of the driving wheel; V is the walking speed. Power supply unit: the system adopts 24 V battery power supply, and the power supply is stabilized by DC24 v-dc24 V switching power supply to ensure the stability of power supply voltage of control system.
Safety unit: ultrasonic radar is installed at the front of AGV, which is used to detect obstacles on the path. An emergency stop button is installed at the front and rear of the AGV to stop the vehicle in case of emergency. When an obstacle is detected, the AGV will decelerate and stop at the specified distance from the obstacle. When the obstacle disappears, the AGV will resume operation automatically. At the same time, a mechanical safety touch device is installed at the front of AGV. When the safety touch device is impacted or squeezed, it is equivalent to pressing the emergency stop button to guide the AGV automatically Stop operation immediately to ensure the safety of staff and production equipment.
Communication unit: the Ethernet equipment in AGV control system of AGV of AGV is connected to LAN by network switch. Data exchange between control devices is based on network to form a single vehicle communication LAN. As shown in Fig. 3, the network topology of communication system of AGV of AGV is shown in Figure 3. The network communication of control system adopts two modes, namely Ethernet and RS485 communication.
The function of Ethernet based network communication is to connect the AGV vehicle control system of a single AGV to the wireless network outside the vehicle and interact with the upper computer on the server side through the vehicle wireless client The current position and posture, all kinds of operation information are uploaded to the upper computer, so that the production scheduling personnel and field staff can monitor the overall running state of the vehicle, and can directly send the task instructions to the vehicle through wireless communication through the server-side host computer. This topic uses Wi Fi communication method to complete the communication between the AGV of the automatic guidance robot and the upper computer of the server.
The AGV master controller of AGV based on RS485 network communication connects motor driver, RFID reader, magnetic navigation sensor, inertial navigation and safety touch detection sensor to RS485 bus, and reads data and sends instructions to all devices through polling. The main controller directly controls the driver through 485 bus, and sends the target steering angle and target running speed to the motor driver to drive the walking motor and steering motor. The RFID reader reads the RFID tag arranged on the running path and transmits the data stored in the tag to the main controller through 485 bus, which is used to confirm the current position of the vehicle. The magnetic navigation sensor is used to measure the deviation position of the ground magnetic nail relative to the sensor. The deviation data is transmitted to the main controller through 485 bus. After the master controller calculates and processes the data, it obtains the actual position deviation and attitude deviation of AGV body of AGV, and compensates the error of differential GPS navigation system itself. Different from the traditional I / O signal transmission mode, the obstacle detection sensor sends the detected obstacle information to the main controller through 485 bus, including the distance between the obstacle and the vehicle body, the size of the obstacle, etc. according to these information, the master controller makes judgment and sends instructions to the electric driver to slow down the AGV or stop the AGV in an emergency.
3. Automatic guidance control method
The navigation system collects and processes the environmental data, determines the current position and attitude of AGV, and sends the deviation from the target position and attitude back to the control system, which is used for the control system to give instructions to the drive system and adjust the position and attitude deviation to ensure the navigation accuracy. This design adopts the hybrid guidance mode of magnetic nail navigation and inertial navigation.
Magnetic nail navigation: similar to the traditional magnetic tape guidance, the magnetic navigation sensor is used to detect the deviation of AGV relative to the magnetic nail to adjust the position and posture of the automatic guidance robot. The difference between them is that the tape is continuous guidance, while the magnetic nail guidance is discontinuous. The path between the magnetic nail and the magnetic nail is a "blind area". In order to ensure that the AGV of the automatic guidance robot walks along the predetermined magnetic nail path, the walking distance of the walking wheel is recorded in real time when walking in the "blind area", the change of the steering angle is calculated, and the current position of the AGV is calculated to adjust the deviation. However, in the process of AGV walking, the wheels of AGV sometimes slip, which will lead to the deviation of the position calculation results, resulting in the low safety and reliability of AGV.
Inertial navigation: using the acceleration and angular velocity integral, the running speed and heading angle of AGV are obtained, and then the current actual position of AGV of AGV is calculated, and compared with the planned path, the actual deviation of current position, namely navigation deviation, is obtained, and then the deviation is corrected. However, due to the performance difference of inertial navigation module, there are different degrees of drift error and noise, and the drift error and noise will accumulate with the increase of time, which will lead to the AGV walking off track and unable to reach the next magnetic nail position.
The lack of magnetic navigation and automatic navigation of the robot can be eliminated by the combined navigation of magnetic navigation and nail. Magnetic nail navigation is used to compensate the drift error and noise of inertial navigation module. The position coordinate of the magnetic nail on the ground is absolute, and the position of the magnetic navigation sensor installed in front of and behind the AGV body of the AGV is absolute. At this time, the current AGV can be calculated by detecting the deviation distance of the magnetic nail on the ground by the magnetic navigation sensor The position and attitude of AGV is used to compensate the accumulated error of inertial navigation module, which improves the operation accuracy of AGV and increases the reliability and safety of AGV.
As shown in Fig. 4 is the schematic diagram of navigation deviation calculation. The direction of the AGV body center line is the walking direction, and the angle α between the AGV and the planned path is the steering angle error, which is the navigation deviation. The calculation of the navigation deviation is shown in formula (1).
It can be seen from Figure 4 that the car body deviates from the planned route to the right
(DG-AE)> 0 (2)
Then there are
If the car body deviates from the planned route to the left, then
(DG-AE)＜ 0 (4)
In the same way
α< 0 （5） It is concluded from (3) and (4) that the AGV calculates the change of α in real time during the walking process. As long as the navigation deviation α is not equal to 0, the main controller PLC will immediately send instructions to control the steering motor action, change the walking direction, and make the AGV return to the predetermined planning path.
4 control system software design
4.1 main program design
As shown in Fig. 5 is the main program flow chart of AGV. After the AGV of AGV is powered on, the bus equipment communication status is detected, including RS485 communication equipment, such as motor driver, RFID reader, magnetic navigation sensor, inertial navigation sensor and safety touch detection device. RS232 communication equipment is differential GPS navigation system and Ethernet communication equipment is wireless client. If any abnormality is found in the detection, the control system will give an alarm. Then enter the state detection of safety device. If it is found that the emergency stop button is pressed, an obstacle is detected or the safety touch detection device acts, the control system will send an alarm and stop. The next step is to calibrate the zero point of the steering motor. Since the encoder of the steering motor uses the incremental encoder, it needs to calibrate the zero point after power on.
4.2 automatic seeker program design
As shown in Fig. 6, the subprogram design flow chart of AGV is shown.
Receiving and parsing task: according to the received information of the starting site and the target site, the path planning is carried out and the map path sequence is generated. According to the path sequence, the AGV of the robot is automatically guided to run.
Confirm the current vehicle position and attitude: receive the longitude, latitude, heading angle and other position and attitude information of the current vehicle sent back by the differential GPS navigation system, and set them as the current position and attitude.
Comparison of vehicle real-time pose and target pose deviation: in the receiving and analyzing task stage, the map path sequence has been obtained, in which the specific position and attitude coordinates of the passing station are included. The position and attitude deviation is calculated by comparing the real-time position and attitude with the information of the next station.
Control the driving motor action to eliminate the deviation: according to the position deviation of the previous step, control the drive motor action to eliminate the deviation.
Acquisition of the ground magnetic nail position and compensation correction of real-time position and attitude: because of the error of differential GPS navigation data, and the magnetic nail fixed on the ground, its position relative to the ground is regarded as absolute coordinates, the AGV body front and rear are installed with a magnetic navigation sensor, its installation position is also absolute coordinates relative to the AGV body. When two magnetic navigation sensors detect the magnetic nail at the same time, the position and attitude deviation of AGV body position relative to the ground magnetic nail coordinate data can be calculated according to the geometric relationship, which can be used to compensate the error of differential GPS navigation data.
Safety device status detection: detect the status of emergency stop button, obstacle detection and safety touch safety devices. In case of any abnormality, AGV will reduce the speed and send out alarm or emergency stop.
5 system man machine interface
The man-machine interface of the system mainly includes start-up interface, navigation setting interface, driving system setting interface, safety touch interface, single task mode setting interface, continuous task setting interface, system setting interface, operation record viewing interface and alarm interface. The data of operation record viewing interface and alarm interface can be retained for one year. The interface is rich in content and easy to operate, as shown in Figure 7 and figure 8.
Aiming at the application of the robot in practice, the control system of the automatic guidance robot based on the combination of magnetic nail and inertial navigation is developed, and the overall design of the whole control system is introduced in detail. In hardware aspect, the function and composition of control system are introduced, the function of each functional unit and the operation strategy meeting the requirements of automatic guidance of control system are described in detail. The construction and operation strategy of communication are emphasized. In the aspect of software, the main program design flow and the automatic pilot subprogram design flow of the whole control system are introduced, and the detailed description is given. At the same time, the calculation method of eliminating navigation deviation in walking process is introduced, which provides the basis for further study of path planning, which has certain practical significance.
 Chang Hao, guzhenchao, Douyan, etc. Design and research of intelligent industrial handling robot [j]. Introduction to scientific and technological innovation, 2020 (15): 69-71
 Jianghanrong, Qian Xiaoming. The extraction of the center line of AGV road marking line based on HSI image segmentation [j]. Science and technology and engineering, 2011,11 (35): 8757-8760
 Wang yunyun, Wang Lianming, guanjianwen, et al. The design of the indoor transportation automatic guided vehicle system based on robot operating system [j]. Science and technology and engineering, 2020,20 (19): 7742-7749
 Wang Peisong, Ma Shiqiang, Wu poisley, et al. Design and Application Research of intelligent object-taking robot [j]. Introduction to scientific and technological innovation, 2020 (15): 88-89
 Zhangyanyan. Design and path planning of automatic guided carrier [d]. Xi'an: Xi'an University of science and technology, 2016