Merge pull request WISDEM#18 from dzalkind/simulink

Simulink

.gitignore

@@ -42,3 +42,11 @@ _build
 .pytest_cache
 examples/cp_ct_cq_lut.p
 
+# Files Generated in Examples
+Examples/DISCON.IN
+Examples/*.p
+
+# Matlab Stuff
+Matlab_Toolbox/*.slxc
+*.mat
+

Examples/.gitignore

@@ -0,0 +1,2 @@
+Cp_Ct_Cq.*.txt
+*.p

Examples/example_11.py

@@ -0,0 +1,83 @@
+'''
+----------- Example_11 --------------
+Load a turbine, tune a controller, export linear model
+-------------------------------------
+
+In this example:
+  - Load a turbine from OpenFAST
+  - Tune a controller
+  - Use tuning parameters to export linear model
+
+'''
+# Python Modules
+import yaml
+# ROSCO toolbox modules 
+from ROSCO_toolbox import controller as ROSCO_controller
+from ROSCO_toolbox import turbine as ROSCO_turbine
+from ROSCO_toolbox import sim as ROSCO_sim
+from ROSCO_toolbox import utilities as ROSCO_utilities
+
+import numpy as np
+
+# Load yaml file 
+parameter_filename = '/Users/dzalkind/Tools/ROSCO_toolbox/Tune_Cases/IEA15MW.yaml'
+inps = yaml.safe_load(open(parameter_filename))
+path_params         = inps['path_params']
+turbine_params      = inps['turbine_params']
+controller_params   = inps['controller_params']
+
+# Linear file output
+linmod_filename     = '/Users/dzalkind/Tools/matlab-toolbox/Play/IEA15MW_LinMod.dat'
+
+# Instantiate turbine, controller, and file processing classes
+turbine         = ROSCO_turbine.Turbine(turbine_params)
+controller      = ROSCO_controller.Controller(controller_params)
+file_processing = ROSCO_utilities.FileProcessing()
+fast_io         = ROSCO_utilities.FAST_IO()
+
+# Load turbine data from OpenFAST and rotor performance text file
+turbine.load_from_fast(path_params['FAST_InputFile'],path_params['FAST_directory'],dev_branch=True,rot_source='txt',txt_filename=path_params['rotor_performance_filename'])
+
+# Tune controller 
+controller.tune_controller(turbine)
+
+# Write Linear model parameters to text file for matlab processing
+# Add to ROSCO_utilities.FileProcessing() when finished
+
+print('Writing linear models to text file: ' + linmod_filename)
+
+# extend gain schedule
+
+pc_br = np.zeros(len(controller.v_below_rated))
+pc_Kp = np.concatenate((pc_br,controller.pc_gain_schedule.Kp))
+pc_Ki = np.concatenate((pc_br,controller.pc_gain_schedule.Ki))
+
+vs_ar = np.zeros(len(controller.pc_gain_schedule.Kp))
+vs_Kp = np.concatenate((controller.vs_gain_schedule.Kp,vs_ar))
+vs_Ki = np.concatenate((controller.vs_gain_schedule.Ki,vs_ar))
+
+with open(linmod_filename,'w') as f:
+    f.write('{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\t{:12}\n'.\
+        format('WindSpeed','A_om','b_theta','b_tau','b_wind','pc_Kp','pc_Ki','vs_Kp','vs_Ki','Pi_omega','Pi_theta','Pi_wind'))
+
+    for v,A,B_beta,B_tau,B_wind,pc_Kp,pc_Ki,vs_Kp,vs_Ki,Pi_omega,Pi_beta,Pi_wind in zip(controller.v,controller.A,controller.B_beta,controller.B_tau, controller.B_wind, \
+        pc_Kp, pc_Ki, vs_Kp, vs_Ki, \
+        controller.Pi_omega, controller.Pi_beta, controller.Pi_wind):
+        f.write('{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\t{:12.4e}\n'\
+            .format(v,A,B_beta,B_tau,B_wind,pc_Kp,pc_Ki,vs_Kp,vs_Ki,Pi_omega,Pi_beta,Pi_wind))
+
+print('Tower Height = {} m'.format(turbine.hubHt))
+print('Platform Freq. = {} rad/s'.format(turbine.ptfm_freq))
+
+
+# Write parameter input file
+# param_file = 'DISCON.IN'   # This must be named DISCON.IN to be seen by the compiled controller binary. 
+# file_processing.write_DISCON(turbine,controller,param_file=param_file, txt_filename=path_params['rotor_performance_filename'])
+
+# Run OpenFAST
+# --- May need to change fastcall if you use a non-standard command to call openfast
+# fast_io.run_openfast(path_params['FAST_directory'], fastcall='openfast', fastfile=path_params['FAST_InputFile'],chdir=True)
+
+
+
+

Matlab_Toolbox/Pl_FastPlots.m

@@ -22,7 +22,7 @@ function Pl_FastPlots(varargin)
 plsw.PD = 1;                    % PD, Primary Dynamics
 plsw.RO = 0;                    % RO, Rotor Performance Outputs
 plsw.Fl1 = 0;                   % Fl1, Basic Floating Parameters
-plsw.AF = 1;                    % All Floating Parameters
+plsw.AF = 0;                    % All Floating Parameters
 plsw.Twr = 0;                   % Twr, Turbine params with Twr Motions
 plsw.Rand = 0;                  % Some random metrics I care about now
 cases = fieldnames(plsw);

Matlab_Toolbox/README.md

@@ -0,0 +1,17 @@
+# Simulink Implementation of ROSCO
+A simulink version of ROSCO has been created for rapid prototyping of new control ideas. The results do not exactly match results from the .DLL version of ROSCO due to the way that Simulink handles initial conditions. These differences change the wind speed estimate slightly and propogate to differences in the torque control and pitch saturation. The following modules in ROSCO have been implemented in Simulink:
+  - TSR tracking torque control
+  - PI gain-scheduled pitch control
+  - Setpoint smoothing control
+  - Extended Kalman Filter wind speed estimator
+  - Pitch Saturation
+  - Floating feedback control
+
+The modules not currently implemented include:
+  - k\omega^2 torque control
+  - Individual pitch control
+  - Shutdown control
+  - Flap control
+
+`runFAST.m` can be used to load the ROSCO parameters from a .IN file, using `load_ROSCO_params.m` and a more detailed version can be found in the `matlab-toolbox` [repository](https://github.com/dzalkind/matlab-toolbox/tree/master/Simulations).
+