Temperature Control Loop Analyzer (TeCLA) Software F. Burzagli - S. De Palo - G. Santangelo (Alenia Spazio) Fburzagl@to.alespazio.it Foreword A typical feature of an active loop thermal system is to guarantee a tight temperature range to external and/or internal users through a stable and fast control dynamics. This goal is obtained through a compromise between - stability ( bullet-proof, relatively slow system) - performance (reactive, potentially nervous system) This compromise is not always possible.
Standard Temperature Control Configuration Loop Heat Load Controller α 3-way Valve Temperature Sensor HX Regenerative Temperature Control Configuration Temperature Sensor HX 3-way Valve α Controller
Approaches to Stability Analysis Method Pros Cons Experimental Testing Numerical Simulation Theoretical Analysis Stability directly assessed for the tested configurations No test set-up costs Utilization of standard simulators Extensive investigation of all the possible loop configurations Worst case identification (parametric) Possible use of robustness indicators Usually expensive Huge number of scenarios to be investigated Stability not guaranteed in other configurations than the tested ones Huge number of scenarios to be investigated Stability not guaranteed in other configurations than the tested ones Modeling uncertainties Simplified representation of the loop Approaches to Stability Analysis (cont d) A detailed analysis of the loop stability and performance has to consider the following points: investigation of stability based on automatic control theory (linear and/or nonlinear) [theoretical analysis] choice of optimal and robust control law coefficients [numerical simulation and theoretical analysis] identification of operational worst case scenarios [theoretical analysis] simulation of transients leading to worst case scenarios (verification of stability regions) [numerical simulation] optional testing campaign focused on worst case operational scenarios (verification of stability regions and system performance) [experimental testing]
Node 2/Columbus Testing Experience Examples of the experimental activity are the testing campaign on Node 2 Breadboard and on Columbus Water Loop Step IV (both carried out in the period July-August 1998). Some limits of the trial and error methods were highlighted: the stability/instability assessment of the system was limited only to the tested cases; various sets of control coefficients were tested to investigate the influence on the system stability; evidence about the reliable identification of the worst case scenario was not obtained; TeCLA Code Temperature Control Loop Analyser (TeCLA) was developed by Alenia to perform stability analysis on Node 2 layout and software. The main features of the code are: The loop layout can be interactively inserted by the user 3-way and heat exchanger hydraulic and thermal data are interpolated by means of neural network the worst case scenario is found inside a range of operational parameters given by the user (and derived from requirements) stability regions in terms of controller gains are evaluated stability regions evaluated on the basis of Nyquist (linear) and circle (nonlinear) criteria robustness indicators (vector, phase or gain margin) are given for each pair inside the stable zone
TeCLA Code - graphic user interface TeCLA Code - stability region screen shot KP KD
Application to Node 2 LT and MT loops (iso-phase margin curves - standard configuration) LTL Standard Configuration - Phase Margin Curves MTL - Worst Scenario - Phase Margin Curves 1.4 5.5 1.2 1.8.6 PM=2 PM=1 no margin 5 4.5 4 3.5 3 2.5 2 PM=2 PM=1 no margin.4.2 PM=3 PM=4 1.5 1.5 PM=4 PM=3-1 1 2 3 4 5 6-2 2 4 6 8 1 12 14 Application to Node 2 LT loop (iso-phase margin curves - regenerative configuration) SPCU Regenerative Configuration - Phase Margin Curves.12.1.8 PM=15 PM=1 PM=5 no margin.6.4 PM=2.2 -.5.5 1 1.5 2 2.5 3 3.5
SINDA-Fluint Simulations vs. TeCLA predictions 2.5 LTL - Standard Configuration - Stability Region [ Q =11.2 kw - T =37 F ] NH3 1 LTL Standard Configuration - Transient Simulation 9 2 8 1.5 1 selected gains Valve position [%] 7 6 5 4 3.5-2 2 4 6 8 1 12 2 1 STABLE CASE 5 1 15 2 25 3 35 4 45 Time [sec] Heat variation: ± 1 kw (max injected power: 11.2 kw) Ammonia T: 37 F - KP = 1.5 - KD = 2 - KI =.1 SINDA-Fluint Simulations vs. TeCLA predictions (cont d) 1.8 LTL - Standard Configuration - Stability Region [ Q =11.2 kw - T =34 F ] NH3 1 LTL Standard Configuration - Transient Simulation 1.5 1.2.9.6 selected gains Valve position [%] 9 8 7 6 5 4 3.3-2 -1 1 2 3 4 5 6 7 8 2 1 UNSTABLE CASE 2 4 6 8 1 12 14 16 18 Time [sec] Heat variation: + 1 kw (max injected power: 11.2 kw) Ammonia T: 34 F - KP = 1.5 - KD = 2 - KI =.1
TeCLA ongoing development The software is being updated in order to: cover a wider range of layout configurations (1% completed, debugging ongoing) cover a wider range of control laws (5% completed) utilize a wider spectra of stability and robustness indicators (2% completed) TeCLA will be structured in two separate units: the solver (ANSI C) runs on Unix platform Intranet/Internet access to graphic pre- and post-processing (Java) Development schedule: mid 22 TeCLA Architecture (ongoing development) Scripting (Java) ANSI C Scripting (Java) Graphic Input I/F Input File TeCLA solver Output File Graphic Post-Processor Graphic Neural I/F Neural DB Neural Network Error Log Graphic Error Report C++ Client Server Client
Future Work Coefficient optimization: Insert an inverse Laplace transform routine to provide information on the system performance inside the stable region Mission control utilization: Link of the software to real-time flight telemetry to analyze off-nominal scenarios and verify the stability of the implemented control law parameters Extension to pressure and mass flow rate controls Software validation: In the framework of Node 2 Testing activity at integration level (if possible) NOTE: once the Java I/F has been assessed for TeCLA, its use can be extended to develop connection tools between mission control telemetry and simulation software (SINDA-FLUINT, ESATAN-FHTS) to analyze flight transient with considerable time saving.