Hi all,
I am currently working on a software in Automation Studio that is supposed to detect when effort is inputted by a user. This is done via a sensor and current transducer across my motor. When no effort is inputted, I should get a certain value, and when effort is applied, there is a drop in the current value. There are a few settings/metrics in the device that can be changed. I have collected data and created formulas that should predict, based on the settings/metrics what the sensor value should be, and in my program, if the sensor value is below this value, "effort" will go high.
However I have been having one issue that has been driving me insane. Due to current decay in the motor over time from overcoming inertia and from "warming up", the "baseline" tends to shift around time to time. It is still a somewhat predictable model , so in context of my formula, the coefficients for each variable is accurate, but the y- intercept is off.
All this to say, is there any thoughts or ideas on how to identify where the baseline is, and account for the shift/decay in my software?
I am currently working on a software in Automation Studio that is supposed to detect when effort is inputted by a user. This is done via a sensor and current transducer across my motor. When no effort is inputted, I should get a certain value, and when effort is applied, there is a drop in the current value. There are a few settings/metrics in the device that can be changed. I have collected data and created formulas that should predict, based on the settings/metrics what the sensor value should be, and in my program, if the sensor value is below this value, "effort" will go high.
However I have been having one issue that has been driving me insane. Due to current decay in the motor over time from overcoming inertia and from "warming up", the "baseline" tends to shift around time to time. It is still a somewhat predictable model , so in context of my formula, the coefficients for each variable is accurate, but the y- intercept is off.
All this to say, is there any thoughts or ideas on how to identify where the baseline is, and account for the shift/decay in my software?