An efficient speed control method for two mass drive systems using ADALINE

An industrial two mass system is composed of a motor which is connected to the load machine through a shaft. Sometimes the shaft may be stiff, but in applications such as rolling mill drives, robotic arms, conveyor belts etc., the shaft is elastic. These elastic joints are susceptible to angular vibrations called torsional vibrations that limit the performance of industrial drives. Due to this the motor speed may be different from the load speed. To solve this problem speed controllers are used which will adjust the load speed and the motor speed with that of the reference speed. Fuzzy logic controllers and PI controllers were used earlier. The limitation of these controllers is the lack of analytical tuning. In recent years, neural network have been applied for the control of two mass systems. Adaptive Linear Neuron (ADALINE) based speed controller with an efficient adaptation algorithm can be used to stabilize the two mass drive systems. The control signal to be applied to the two mass system is generated by the ADALINE controller which updates its weight value on demand. The design of modules can be done in Xilinx System Generator. System generator enables the use of MATLAB/Simulink for simulating the designed modules. It also enables the user to create user defined blocks. The modules that are designed in System Generator can be implemented in FPGA. The work is further extended by modifying the linear activation function present in the ADALINE module by non linear activation function.