In order to solve the problem of low working efficiency of the electrical control system of the traditional picking manipulator, the application research of PLC in the manipulator control was proposed. Based on PLC controller and computer image processing technology, this paper proposes a control system optimization algorithm based on particle swarm optimization. Through the recognition and location of the fruit image to be picked, the best picking sequence is drawn out by using the particle swarm algorithm, which shortens the end moving distance of the picking manipulator and improves the operating efficiency. In order to verify the feasibility of the scheme, taking the real green pepper picking environment as the object of image acquisition, the distance and picking accuracy of the manipulator moving end with and without particle swarm optimization algorithm were tested. The experimental results show that the moving distance of 19.5m/s with 50 green peppers using PSO algorithm is significantly lower than that of 29.6m/s without PSO algorithm, and the picking efficiency is high. This is because the adoption of particle swarm optimization algorithm can help the picking manipulator to reasonably plan the picking sequence and complete the task with the shortest distance, thus improving the work efficiency.