Abstract: It is the current general energy-saving control method that a matching relationship between coal volume and belt speed is used to control the belt speed. However，the influences of the sensor installation position and data processing speed will cause the detection signal to lag behind the real-time coal information. When the coal volume suddenly increases, the belt conveyor cannot accelerate in time to cause the occurrence of coal accumulation and overflow. In order to solve the above problems, a new energy-saving control method is proposed in this paper. The state observer of the belt conveyor system is established to estimate the tension value of each section on the conveyor belt. And then，combining the change rate of the belt tension value at the head drive roller with the coal volume information tested by the electronic belt scale in real time，an new energy-saving system is established by applying fuzzy reasoning and decision-making to control the belt conveyor accelerate when the coal volume increase. Based on this, basing on the linear steady-state system model of belt conveyor start-up process, a time-varying system dynamics model of belt conveyor based on load change is established, and the hardware in loop test is performed by using NI CompactRIO-9038 real-time controller, Simulink and LabVIEW software .The verification results show that:when the coal volume does not fluctuate greatly, the conveyor maintains a constant speed, indicating that the system has strong anti-interference ability;in the case of sudden change of coal volume, the speed-up effect is obvious with the auxiliary help of the tension change rate, and these issues of the coal accumulation and overflow can be solved.