Hokkaido University Research Profiles

Japanese
Manufacturing Technology

Nonlinear Compensator That Can Be Implemented Without Sensors

Nonlinear compensator that can easily be added to PID control systems

Currently, PID control is used as the main control method in industry, but the PID control technique has a problem that the control accuracy deteriorates due to the influence of nonlinear terms such as friction and gravity. We have proposed a nonlinear compensator that can easily be added to PID controllers.

Content of research

Digital acceleration control (DAC) is a robust control technique for systems with nonlinear terms and modeling errors that are difficult to model. DAC is a very effective controller, but it cannot perform position control by itself because it controls the target acceleration value. Therefore, we have combined DAC with a general PID control system. This PID-DAC combined control system allows both robust position control and acceleration control. In addition, as a new nonlinear compensator that can easily be added to PID controllers without sensors, we propose two controllers: the PID-DA0 control system, which sets the target acceleration value of the control object to zero, and the PID-DJ0 control system, which sets the target acceleration value to zero. Both controllers can easily be added to existing PID controllers without additional sensors, so they have the great advantage of improving system performance sensorless.

Potential for social implementation

  • ・Since it is possible to add a sensorless nonlinear compensation function to a system that is currently PID-controlled, we believe that this method has significant application possibilities, including the position control of manipulators.

Appealing points to industry and local governments

PID control is known as a widely-used control technique, and 97% or more of companies in the industry use it. Since this method can easily be added to a PID controller to compensate for nonlinear terms such as friction and gravity, we believe that it has a large potential market. In particular, it is expected to be widely applied to the control of mechanical systems, including industrial manipulators.

2022/5/27Released