Abstract: In order to realize the goal of robotization and unmanned operation，and to improve the efficiency and intelli⁃ gence degree of cutting coal and gangue with roadheader，a control method and system of intelligent cutting on coal and gangue with roadheader has been put forward. The principle and hardware components of the system are described in detail. The intelligent gangue cutting control system of roadheader uses a variety of sensor information collected in real time as control variables，mainly including cutting motor current I，cylinder pressure p and cutting arm vibration acceleration acc ，to achieve the identification of gangue in different cutting states of the roadheader and drive the intelligent cutting of the roadheader’s cutting arm. The hardware part of the system consists of a sensing and detection sys⁃ tem，an on⁃board master control system and a system actuator，and a remote monitoring platform. According to the elec⁃ trical standards of coal industry，the cutting control strategies of general working conditions and special working condi⁃ tions are constructed respectively. Under the general working conditions，the BP neural network controller optimized by the improved particle swarm optimization algorithm based on a variety of sensor information is used to realize cutting load identification，improve the accuracy of coal and gangue identification of roadheader，and realize the adaptive cut⁃ ting arm swing velocity. In the special working conditions，the control logics of hard spot identification and stuffy car prevention are constructed respectively，and the judgment and treatment of special working conditions are realized by Automation Studio software programming. Taking EBZ135 roadheader as an example， the verification experiment is carried out. The experimental results show that the adaptive adjustment time of the swing velocity of the cutting arm is about 0.8 s and there is no overshoot，which has good robustness. The experimental results of hard point identifica⁃ tion control and stuffy car prevention control show that the system is accurate and fast for the special working condi⁃ tions recognition，and the designed control strategy has a high reliability and control precision.