Aiming at the control problem of the automatic guided vehicle, the kinematics model of the two-wheel differential automatic guided vehicle is established firstly. According to the kinematics model, a fuzzy controller suitable for the model is designed by using the theory of fuzzy control.
The digital simulation results show that the designed fuzzy controller has good control effect on the two-wheel differential automatic guided vehicle. Automated Guided Vehicle (AGV), as a mobile robot, is widely used in modern manufacturing system and automatic warehousing system for automated logistics transportation. The control method of AGV has always been the focus of scholars’research.
Fuzzy control is independent of precise mathematical model and has good robustness.
It reduces the complexity of AGV control system, has fast information identification speed and high flexibility of path setting. It has become an important direction of AGV development. The AGV studied in this paper is a two-wheel differential system. The guidance information obtained by AGV through the front and rear magnetic guidance sensors, i.
e. the displacement and angle deviation between AGV and the guide magnetic strip, tracks the path of the laid magnetic strip. Therefore, the algorithm based on the global kinematics model is difficult to apply [1-2]. The kinematics model of AGV is shown in Figure 1. AGV is a highly non-linear, strong coupling and time-varying dynamic system. It is difficult to establish its accurate dynamic model and control it. Fuzzy control method does not depend on mathematical model and is especially effective for the control of non-linear systems. Fuzzy control establishes fuzzy rules based on expert experience, which provides an effective method for AGV control.
Magnetically guided AGV is mainly based on magnetic guided sensor for path tracking to complete its autonomous driving function. The position and attitude deviation between AGV and target path is judged by detecting the magnetic strips laid on the ground, and the switch level signal is output [5-6].
Because the output signal of magnetic guidance sensor is easy to collect and process, the control method of magnetic guidance AGV is simple and reliable compared with visual guidance and laser navigation. When the AGV is running, the magnetic guidance sensor installed in the front and back end of the AGV collects the magnetic strip signal to determine the position and attitude deviation of the AGV.
The distribution of the induction points of the magnetic guidance sensor is shown in Figure 2.
Using optocoupler, the feedback is set to be 1 when the signal is collected by the magnetic sensor, thermostatic element otherwise it is 0. If AGV deviates from orbit on the left, set ed to be negative, and if AGV deviates from orbit on the right, set ed to be positive. The effective detection range of a single magnetic sensor of a magnetic guidance sensor is a circle with a radius of 15 mm and the magnetic sensor is the center of the circle. Number each magnetic sensor as shown in Figure 2.
In formula (6), for example, NB is only 6 detected signals, NM is 5 and 6 simultaneously detected signals, NS is 4 and 5 simultaneously detected signals, ZR is 3 and 4 simultaneously detected signals, PS is 2 and 3 simultaneously detected signals, PM is 1 and 2 simultaneously detected signals, PB is only 1 detected signals.
To verify the effectiveness of the designed fuzzy controller, a large number of numerical simulations [7-8] are carried out under different velocity V and different pose deviations. When the initial path deviation is eTheta=13 degrees, ed=0 and eTheta=18 degrees, ed=0, respectively, the numerical simulation results are shown in Fig. 4 (a) and Fig. 4 (b). The velocity of AGV in Fig. 4 (a) is 0.2m/s, and that of AGV in Fig. 4 (b) is 0.5m/s. The simulation results show that the fuzzy controller designed in this paper can correct the position and attitude deviation quickly. Aiming at the path tracking problem of AGV, the kinematics model of two-wheel differential AGV is established, and a fuzzy controller for this model is designed. The digital simulation results show that the fuzzy control method selected in this paper has better control performance for this AGV.