| Ball mill is the key equipment in the process of re-smashing after the material breaks,and ball mill's load is an important parameter of the grinding process. Nowadays single-factor method or two-factor method is adopted to measure the external response and the inner load hasn't been forecasted precisely. The ball mill's load can't be detected precisely by these methods,and ball mill barely work under optimizing state.Referencing on domestic and overseas relevant study, the three-factor method was put forward and used to detect external three response: noise, vibration and active power. According to frequency analysis, noise and vibration's power spectrums are estimated. The disposed response signals are basis for forecasting internal load parameters. Using RBF neural network, the relation model between internal load and external response was funded with noise, vibration and active power as input signals, and ratio of ore and ball, proportion of ball's volume to ball mill's volume and grinded ore's concentration as output signals. On swatch date, the structure, weights and threshold values of the model was gained by training. The simulation results show that the model can forecast the ball mill's load precisely. Internal parameter (ratio of ore and ball, proportion of ball's volume to ball mill's volume and grinded ore's concentration ) can be forecasted successfully by these means.The embedded technology is an emerging technology.When Embedded systems connect to the Internet, it is convenient cheaply to send information to almost any place in the world. when Ball mill works,it sends out the noise,The environment is bad.When the ball mill load examination system turns on Internet, the human then may observe the movement condition and each kind of parameter of the scene equipment through Internet.The human does not need to arrive personally the scene. Through network long-distance accusation object monitoring is realizated . This examination system mainly applicates binuclear CPU - - chip TMS320VC5471.The ARM7TDMI achievement advocates CPU to be responsible for the complete machine the control, the communication management and the man-machine connection function. The real-time operating system -μC/OS-Ⅱis transplanted to the ARM7TDMI nucleus. Then TCP/ IP network protocol is designed on the real-time operating system.DSP took from CPU is signal gathering and the data processing core. It completes the sound, the vibration and active power signal gathering and the data processing function and realizes the ball mill load survey and the ball mill running status online examination. |