(1) Guidance and positioning technology. As the core part of AGV technology research, the quality of guidance and positioning technology will directly affect the performance stability, automation degree and application practicability of AGV.

(2) Path planning and task scheduling technology. First, route planning. Driving path planning is to solve the problem of how to get to the starting point of AGV. At present, a large number of artificial intelligence algorithms have been used in AGV path planning at home and abroad, such as ant colony algorithm, genetic algorithm, graph theory, virtual force method, neural network and a * algorithm.

Second, job scheduling. Job task scheduling refers to processing tasks according to the request of current jobs, including sorting tasks based on certain rules and arranging appropriate AGV processing tasks. It is necessary to consider the task execution times, power supply time, work and idle time of each AGV, so as to achieve the rational application and optimal allocation of resources.

Third, multi machine coordination. Multi machine coordination refers to how to effectively use multiple AGVs to complete a complex task together, and solve a series of problems such as system conflict, resource competition and deadlock. At present, the commonly used multi-machine coordination methods include distributed coordination control method, road traffic rule control method, multi-agent theory based control method and multi robot control method based on Petri net theory.

(3) Motion control technology. Different wheel mechanisms and layouts have different steering and control modes. At present, the steering drive modes of AGV include the following two: two wheel differential drive steering mode, that is, two independent driving wheels are fixed in the middle of the vehicle body in parallel, and other free universal wheels are used for supporting. The controller can realize any turning radius by adjusting the speed and steering of the two driving wheels Steering: the steering wheel controls the steering mode, that is, the steering is realized by controlling the yaw angle of the steering wheel, which has the limit of the minimum turning radius.

The control system is composed of a closed-loop system by the feedback of the encoder installed on the drive shaft. At present, the AGV path tracking methods based on the two-wheel differential drive mainly include: PID control method, optimal predictive control method, expert system control method, neural network control method and fuzzy control method.

(4) Information fusion technology. Information fusion refers to the use of Multi-source Information Association combination, fully identify, analyze, estimate and schedule data, complete the task of decision-making and accurate processing of information, and make appropriate estimates of the surrounding environment, war situation, etc. At present, Kalman filtering, Bayesian estimation and D-S evidence reasoning are the most widely used information fusion technologies in the field of guidance. Kalman filtering has a good real-time performance, but it is based on a strict mathematical model. When there is a large modeling error in the guidance model or the system characteristics change, it often leads to filter divergence. In order to improve the robustness and adaptive ability of the filtering algorithm, the adaptive Kalman filtering algorithm, robust filtering algorithm or intelligent filtering (such as fuzzy reasoning, neural network, expert system) method can be studied according to the guidance requirements and characteristics of AGV.