Research On Drive Optimization Of Overhead Crane Based On Ant Colony Algorithm And BP Neural Network

As one of the most important material handling equipment in the modern production process,hoisting machinery is widely used in workshops,docks,ports and construction sites.With the rapid expansion of China’s economic scale,the role of cranes in the modern production process is becoming more and more important.However,in the field of cranes,there is still a lack of speed regulation control of motors,and it is not possible to satisfy some operations that require high control accuracy,such as cranes for nuclear power plants,rotor lifting of aeroengines,automatic path planning for bridge cranes,and so on.In particular,the intelligentization of cranes is the future development trend,and the speed control of the motor is a very important part of its intelligentization.Proportional integral differential control(hereinafter referred to as PID control)is widely used to control motors because of its simple structure,simple implementation and strong stability,but the traditional PID control parameters are fixed,and its control effect is not very ideal.In traditional PID control The lower motor responds slowly and has a large overshoot.The BP neural network is a feed-forward network.The algorithm uses gradient descent to modify the weights and thresholds until the error reaches a minimum,which can achieve online adjustment of parameters according to the current error.However,this algorithm has the problems of easy to fall into local minimum and sensitive to the initial parameter setting.Therefore,the paper proposes a BP neural networkbased on ant colony algorithm to optimize the additional momentum coefficients to achieve online real-time adjustment of PID parameters to optimize PID control performance.The ant colony algorithm is used to search the optimal initial parameter values globally,and then the BP neural network further self-learns to modify the parameters.At the same time,the momentum coefficient is added in the process of modifying the weights to reduce system oscillations.MATLAB simulation results show that the PID controller based on BP neural network tuning based on ant colony algorithm to optimize additional momentum coefficients in this paper has excellent control performance,fast motor response,small overshoot,strong anti-interference ability,and torque fluctuation.Stable,at the same time,it can be applied to automatic rectification,anti-sway and path planning of overhead cranes.It provides a good drive for crane intelligence,helps to improve production efficiency and lifting accuracy,and has certain application value.