Path Planning and Positioning Technology of Cotton Picking Robot in Complex Cotton Field

Abstract

Agricultural robot will become a powerful tool to save a lot of human resources in the future, which can reduce the labor intensity and cost of workers, and improve the efficiency of picking operations. The purpose of this paper is to study the path planning and positioning technology of cotton picking robot based on the complex cotton field environment. Firstly, based on the research of common path planning technology, according to the known degree of the target to the environment, the common path planning algorithms are roughly divided into two categories: global and local planning, and its characteristics, advantages and disadvantages in application are briefly analyzed. The parallel binocular vision measurement system is selected as the positioning tool of the cotton peach to simulate the complex cotton field environment. The experimental results show that after error compensation, the measurement errors in the X and Z directions are within 5mm. The measurement error prediction model for the vision system well expresses the nonlinear model between the original measurement value and the measurement error. The global optimization rate of hybrid particle swarm optimization algorithm is 50%, but the result is the most unstable; the global optimization rate of basic genetic algorithm is 20%, the global optimization rate of multi population genetic algorithm is 70%, the global optimization rate of improved multi population genetic algorithm is 90%, the global optimization rate and speed of improved multi population genetic algorithm are the best of the several algorithms studied in this paper, Therefore, this algorithm is used to optimize the path in practical application.