What's New in ODYSSEE CAE 2024.2

ODYSSEE CAE uses knowledge of a model's time-dependent behavior to:

• Predict the response of the model for parametric surveys.
• Use this prediction as an initial guess for improving the model or to test the model via a reduced-order-modeling (ROM) simulation coupled with optimization.

ODYSSEE CAE includes: 

• ODYSSEE LUNAR: The main GUI of ODYSSEE CAE product and proposes multiple predictions and analyzes. In a few clicks, we can obtain time responses in real-time, optimization and parameter fitting, population studies (uncertainty corridors), sensitivity analysis, and parameters effects also providing animations when applicable. 
• ODYSSEE QUASAR: A solver to make predictive models and data mining analysis. 
• ODYSSEE NOVA: An optimizer to search the optimal variables to minimize an objective function, by respecting constraints. 
• ODYSSEE PULSAR: A tool proposing extraction and replacement processes to handle the input csv data in Lunar format. 

What's New in ODYSSEE CAE 2024.2 Patch

• Bug fixes
  • The optimization process was successfully executed in step 5-Optimization, but the final optimal result and convergence curves were not displayed. This is fixed.
  • For the NSGAII optimizer,
    • When the number of variables was greater than the number of objectives, the calculation was not completed. It is fixed.
    • When an inequality constraint (H<val) was defined, the pareto front displayed points respecting the opposite of the defined constraint (H>val). This is fixed.

What's New in ODYSSEE CAE 2024.2

ODYSSEE Lunar – GUI

• In 2-Data menu, Integration of 2 User Script capabilities with the new check data button located above the Validate Data Configuration button:
  • Check dependences between parameters X to inform you if some parameters have a high correlation. If so, columns are highlighted in the data array of 2-data and it advises you to deactivate one of them.
  • Check duplicate points to inform if there are some identical cases in the database. If so, rows are highlighted in the data array of 2-data.
• In the step 4-interpolation window,
  • Integration of the Compare prediction methods User scripts. We have now two tabs in this configuration window:
    • The first tab to manually select method like in previous versions,
    • The second tab is to configure and execute various compare processes. When it is finished a window appears asking you to apply or not the proposed prediction methods.
  • Separately train datasets: previously, when “do the training step” was on “yes”, the decomposition and/or training part (for neural network) were executed for the whole datasets. As some training can vary in time depending, it is now applied per dataset.
• In sensitivity tab, in the biplot list of parameters’ effect tab, 2 new biplots are added:
  • A parity plot showing the predictive results versus the expected results of the validation cases on the full time. This biplot is visible only if some XVAL - validation cases and YVAL – validation cases have been selected in data management (step 2-Data).
  • A sensitivity ratio plot showing for each cursor position in time, the sensitivity ratio value. This is the ratio of a parameter’s sensitivity compared to the global variation of the response. Please refer to the help manual for more details.
• In optimization, a new method proposing Global optimization with multi objectives and optional constraints named NSGAII has been added in the end of the list. This method is dedicated to plot a front pareto between two defined targets F.
• Enhanced link with ODYSSEE Explore:
  • Explore Data analytic plots automatically loaded with the ODYSSEE CAE project data: when we are clicking on the ODYSSEE Explore icon in the top menu. Data is presented in multiple dashboards, dedicated to analyzing either X and Xval or X with the Y datasets.
  • Files from ODYSSEE CAE used in the dashboards of ODYSSEE Explore are grouped in a new dedicated folders located in:  CAE_ProjectName/odx_XXval  and  CAE_ProjectName/datasetName/odx_XY_ datasetName.
• Bug Fixes
  • In sensitivity: The regression function used to compute the regression coefficient on XY database has been updated to handle the correlated predictors (columns of X input matrix).

We are now using COEFFS = ext("mining", "effectsqr", X, Y, "parameter_effect_")

• • Animation improvements for EnSight format database:

• • With the Fluent-sourced EnSight format, at the end of the animation prediction, we have an error message “not enough memory available to display the animation". It is fixed.
  • The animation prediction was hanging if full database was not selected. It is fixed.
  • The animation prediction of EnSight files coming from Powerflow was not working well. It is fixed.

• • When we were using a partial data from the XVAL file loaded, then the name of the prediction curves in the legend was incorrect. It is fixed.

ODYSSEE Quasar – Solver

• New Statistic function implemented:
  • kde: kernel density estimation (KDE) function corresponds to the kernel smoothing of a finite data sample (matrix X), which output a continuous probability density estimation at each output point. The most common kernel is a Gaussian function. The last argument nbrOutputPoints determines number of equidistant points to compute KDE at.

ext("statistics", "kde", X, nbrOutputPoints)

• • Quantile function gives sample quantiles: zeroth quantile or minimum value (Q0), Quantile (Q1), 2nd quantile or median (Q2), Quantile (Q3), 4th quantile or maximum value (Q4) computed for each column of input matrix X.

ext("statistics", "quantile", X, computeMethod)

• • Histogram function has been documented: This function displays the frequencies of data included in the different bins defined by the user with the variable nbr_bins.

ext("operator","matrix","histogram", X, nbr_bins)

• • effects_withstats function provides more information for a linear regression than just the values of coefficients, based on some assumptions for a linear regression:

• • standard deviation error associated to the coefficient computed.
  • Confident interval for the coefficient, based on provided confidence level alpha.
  • p-value that shows the significance of a variable. p-value determines if obtained effects pf a variable, could be simply due to randomness or chance.
  • sensitivity index (SI) which shows how much the output can change (relatively) by changing independently each input variable. it is normalized based on range of Y.
  • Confidence and Prediction intervals for the linear model.

ext("mining","effects_withstats",X,Y, norm, order, "coeff_output", alpha)

• Modified Unet function has been added. It is an images processing algorithm using a semantic segmentation process that consists of labeling each pixel of an input PNG image with a class. This function based on neural network is using for the learning part the vgg pretrained model and a defined user database of PNG and semantic images (mask). Learning dataset size can be relatively small: starting from tens. The number of classes in the images must be given by the user.

learnModifiedUnet(lr, batchsize, epochs, resize_width, resize_height, num_classes, validationPercent, dataset_path_img, dataset_path_mask, output_model_name)

predictModifiedUnet(New_image_file, resize_width, resize_height, num_classes, model_name, save_file, DisplayOutput)

• Transfer_NN function has been added. This neural network-based classifier predicts the class of an image. It applies a transfer model: learning a new task is done by transferring knowledge from a related task already learned. Based on a learning dataset that contains different classes, Transfer_NN learns classify the new images.

learnTransfer_NN(double learning_rate, int batch_size, int n_epochs, String type, String output_model_name,String[] class_folders_paths)

predictTransfer_NN(String img_path, String output_model_name)

• VAR (Vector Auto Regression) function is an additional method used in time series analysis to model the dynamic relationship between multiple time series variables: Y(t) = function(Y, X1, X2,.., from time t-1 to t-n).

VAR(X, Y, X_toTest, Yinit, regression_order, difference_type, integrator)

• Bug Fixes
  • The regression function used to compute the regression coefficient on XY database has been updated to handle the correlated predictors (columns of X input matrix).

We are now using the following simpler function:

COEFFS = ext("mining", "effectsqr", X, Y, "parameter_effect_")

instead of

COEFFS = ext("mining", "effects", X, Y, 0, ORDER, "parameter_effect_")

ODYSSEE Nova – Optimizer

• NSGAII (Non-dominated Sorting Genetic Algorithm II), is a new global multi-objective evolutionary algorithm, method #19, looking for the best candidates to plot an optimal Front Pareto. To reach that, the points are ranked in level where the front rank contains candidates with a pareto distance is minimum, with no violating constraints.  So, as iterations go on, the algorithm seeks to obtain a single rank with a minimum crowding distance between candidates to plot the optimal Front Pareto.
• Bugs fixed:
  • For NLPQLP and SLSQP, nova was able stay lock at the end of the iteration loops without giving up control.
  • For the DNN function executed under Linux, the function was launched but remained frozen before the “epoch” message. It is fixed.
  • A French message was displayed if the bad path of Nova.exe was used. It is now translated.