| In recent years,with the rapid development of high-tech technologies such as the Internet and artificial intelligence,new technologies and new products of robots have also emerged in large numbers.This not only provides a new breakthrough for the intelligent manufacturing industry,but also improves people's lives from all aspects.The application of robot technology has spread to various fields.Among them,in the construction of a "smart prison" that deeply integrates the Internet,the Internet of Things,artificial intelligence and big data,security robots have made great contributions to improving the intelligent management level of the prison.As a security robot,the night patrol robot is mainly used in indoor scenes such as detention centers and prisons.Through anthropomorphic night patrol monitoring work,it can replace the staff on duty to perform night patrol tasks in the prison,which can reduce the work pressure of the guards on duty and reduce the handling of prison guards.Dangerous,improve the safety supervision level of the prison.This paper designs the hardware system and software control process of the night patrol robot by investigating the field scene of the prison,comprehensively monitoring the indicators,supervising the policing process,etc.At the same time,it studies the compensation algorithm of the robot motion error.First,introduce the actual situation of the prison,conduct a demand analysis,and elicit the working functions of the night patrol robot.On this basis,the hardware system of the night patrol robot is researched and designed,which mainly includes three major modules-chassis,gimbal and power supply.By analyzing the advantages and disadvantages of the omnidirectional mobile chassis,combined with reality,the main body of the chassis uses a three-wheel omnidirectional platform,and selects the appropriate motor and driving device according to the calculation.According to the requirements,the PTZ has designed sub-modules such as lifting mechanism,rotating mechanism,pitching mechanism and telescopic mechanism to adjust the height,angle,level and distance of the vision module,analyze it and complete the selection of the corresponding motor.Combined with the characteristics of long working hours,a power management system is designed,which can monitor the power in real time and charge independently.Secondly,according to the hardware system and functional requirements,design the control methods and processes including the chassis,the gimbal and the power supply.The chassis has established a bus-type control network,which solves the problem of complicated wiring of the traditional mechanical transmission mechanism of the inspection robot,which effectively reduces the interference in communication and reduces the jitter of the chassis.The gimbal is connected to the sub-modules of the gimbal,such as lifting,rotating,and expanding,in turn through the bus,which reduces the complicated requirements of the ports and improves the control accuracy.At the same time,the self-checking and control processes are designed for the PTZ to avoid entanglement.Design the power management process,monitor the power in real time and charge independently.Finally,the error source that caused the displacement error during the robot's movement was analyzed.It was proposed to use a neural network-based compensation coefficient determination method to compensate and simulate and experiment.The motion results at the same command speed before and after compensation were analyzed to verify the proposed The feasibility of the displacement compensation method. |