The robot is a very complex multi-input multioutput nonlinear system. It has time-varying, strong coupling, and nonlinear dynamic characteristics, and its control is very complicated. Due to the inaccuracy of measurement and modeling, coupled with changes in the load and the influence of external disturbances, it is actually impossible to obtain an accurate and complete dynamic model of the robot. We must face the existence of various uncertain factors of the robot. This paper analyzes the real-time communication protocol in the wireless network control system and confirms that the main way to improve the real-time performance of the wireless network control system is to implement the real-time media access control (MAC) protocol. This paper studies robots from the perspective of control and mainly discusses how to use artificial immune algorithms to design robust nonlinear proportion integral derivative (PID) controllers. A nonlinear PID controller is used to replace the classic PID controller. The nonlinear link can be adjusted with the change of the error, so as to achieve the purpose of improving the adaptability and robustness to obtain satisfactory tracking performance. We carried out selective compliance assembly robot arm (SCARA) robot remote control experiment, dual robot following experiment, SCARA and ABB robot collisionless motion planning experiment, and multirobot intelligent collaborative assembly experiment. The experimental results show that the C/S mode remote control system has good practicability and can complete remote tasks; the P2P communication system has good information transmission effects and can realize real-time information sharing between robots; the collision-free motion planning algorithm enables the dual robots to complete obstacle avoidance tasks well in complex operating environments; the functional modules of the system can closely cooperate to complete the tasks in coordination, and the multirobot system has a certain degree of intelligence.

中文翻译：

---

基于无线通信网络的多智能机器人控制系统优化

机器人是一个非常复杂的多输入多输出非线性系统。它具有时变、强耦合和非线性动态特性，其控制非常复杂。由于测量和建模的不准确，再加上载荷的变化和外界扰动的影响，实际上不可能得到一个准确完整的机器人动力学模型。我们必须面对机器人存在的各种不确定因素。本文分析了无线网络控制系统中的实时通信协议，并确认了提高无线网络控制系统实时性能的主要途径是实现实时媒体访问控制(MAC)协议。本文从控制的角度研究机器人，主要讨论如何使用人工免疫算法来设计鲁棒的非线性比例积分微分（PID）控制器。非线性 PID 控制器用于替代经典的 PID 控制器。非线性链路可以随着误差的变化进行调整，从而达到提高适应性和鲁棒性以获得满意跟踪性能的目的。我们开展了选择性顺从装配机器人手臂（SCARA）机器人遥控实验、双机器人跟随实验、SCARA与ABB机器人无碰撞运动规划实验、多机器人智能协同装配实验。实验结果表明，C/S模式远程控制系统具有良好的实用性，能够完成远程任务；P2P通信系统具有良好的信息传输效果，可以实现机器人之间的实时信息共享；无碰撞运动规划算法使双机器人能够在复杂的作业环境中很好地完成避障任务；系统各功能模块可以紧密配合，协同完成任务，多机器人系统具有一定的智能化程度。