Search the Community
Showing results for tags 'model'.
Use Case: It is common in industry to seek to use the behavior of upstream process variables to predict what the behavior of a downstream variable might be minutes, hours or days from the present time. Solution: A traditional predictive modeling workflow can be applied to solve this problem. Identify an appropriate training data set Perform any necessary data cleansing Create a predictive model Evaluate the model fit Improve the model Operationalize the model What differentiates this use case from any other predictive modeling use case is a specific data cleansing step for adjusting signals to remove process lag. 1. Load Data Load your target signal and the relevant upstream signals into the display pane. In this example, the target signal is the product viscosity, measured in an analytical lab based off a sample from a downstream sample point. Three upstream signals: the reactor temperature, reactant conversion, and viscosity modifier flow to the reactor significantly influence the product viscosity measurement and will be used as inputs into the model signal. 2. Identify Training Data Set Identify an appropriate training data set for your regression model. This may involve a longer time window to include variability in product type or seasonality. In this example, we will pan out to 3 months to capture multiple cycles of different product types. With an appropriate training window identified, you can also limit your training data set to a subset of samples present during a particular condition. If this interests you, consult the "advanced options" section of the Prediction Tool KB article for more information. This method is particularly useful if you're wishing to create different models for different modes of operation. 3. Cleanse Signals -- Adjust for Process Lag We can time-shift our upstream signals using either a known constant delay, a known variable delay (like a calculated residence time signal), or an unknown delay of maximum correlation to the target signal. The first two of these options will utilize the .move() function in Formula (or .delay() in earlier versions of Seeq). The latter will utilize the .correlationOffset() function. In this example, we have a known lag of 1.5 hours between the reactor and the product sampling point. We will use the move function with an input scalar of 1.5h, as shown below. The time shift calculation should be applied across all relevant input signals. More information on the different options for time shifting signals using fixed, variable, or calculated offsets is available in this forum post: 4. Cleanse Signals -- Remove signal noise, outliers, abnormal operating data In this example, we apply an agileFilter to each of our time-shifted model input signals. Apply the same technique to each of the model inputs. Note that steps 3 & 4 could have been combined into a single formula. An example of this would be: $reactor_temp.move(1.5h).agileFilter(1min) For guidance on additional cleansing techniques, consult the Interactive Training. 5. Build the Predictive Model Use the Prediction tool panel to create a model of your target signal based on your cleansed, time-adjusted input signals. Ensure your model training window matches the date range that you identified in step 2. You can view the model parameters like coefficients, rSquared, and p-values using the "+ Prediction Model" option. 6. Evaluate the Model Fit Use Scatter Plot view and the model parameters to evaluate the goodness of fit of the model. Switch to a time range outside of your training data set to ensure your model is a good fit for data throughout time. 7. Improve Model (as needed) If the scatter plot indicates a non-linear relationship, test out additional model scales in the prediction tool panel. Consider eliminating variables with p-values higher than your significance level cutoff (frequently 0.05). Add additional variables if relevant. If distinct modes of operation introduce significant signal variability, consider creating a model for each operating mode and stitch the models together into a single model using the splice() function in Formula. 8. Deploy the Model The model should project out into the future by the amount of the process lag between the upstream and target signals.
Scenario: I have created a regression or prediction model for my process but i want to apply that same regression model to another set of signals or a different period of time. This could be helpful for comparing how one piece of equipment is operating when compared against a regression built for another system. It could be used to predict how a system will behave based on how some other similar system behaved. It could be used predict how a system will behave before you have enough information to build that system or run its own model, or any number of applications where we might want to apply a regression or prediction model built on one set of signals to another set of signals. Solution: First off, we have to build a prediction model! In this case i have modeled a Filtration system, predicting filter head loss. I took into account the time the filter has been in service, the raw water turbidity, and the Filter turbidity. Next, if i click on the information icon for my prediction signal, i am able to see the formula that was used to create this prediction. We will need to copy this formula: Next, I need to identify my new signals. In my case, i am going to apply this prediction model to Filter 12. Once i have all of my signals that will be sued to the prediction i can paste my formula into a new formula and make sure all my variables line up (or, alternately and a bit easier, just use the duplicate to formula function from the information screen shown above by clicking on the down arrow next to the duplicate button). Finally, I need to update the PREDICT section of my formula denoted by the .predict() function with the signals i will apply my prediction model to. I can do this by using the search button in the formula tool to add my new signals, and then update the .predict() section of my formula with my new variables, make sure to put them in correctly and in the right order otherwise your model will be off! Finally, i can use my prediction model that i built for Filter 1 on Filter 12 and derive further insight, in this case i might question whether my systems are truly as similar as i think they are or whether there is something causing my model to deviate from what was expected for filter 12. In my case, both prediction models are in red: