1、_ SAE Technical Standards Board Rules provide that: “This report is published by SAE to advance the state of technical and engineering sciences. The use of this report is entirely voluntary, and its applicability and suitability for any particular use, including any patent infringement arising there
2、from, is the sole responsibility of the user.” SAE reviews each technical report at least every five years at which time it may be revised, reaffirmed, stabilized, or cancelled. SAE invites your written comments and suggestions. Copyright 2015 SAE International All rights reserved. No part of this p
3、ublication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. TO PLACE A DOCUMENT ORDER: Tel: 877-606-7323 (inside USA and Canada) Tel: +1 724-776-497
4、0 (outside USA) Fax: 724-776-0790 Email: CustomerServicesae.org SAE WEB ADDRESS: http:/www.sae.org SAE values your input. To provide feedback on this Technical Report, please visit http:/www.sae.org/technical/standards/AIR6326 AEROSPACE INFORMATION REPORT AIR6326 Issued 2015-08 Aircraft Electrical P
5、ower Systems Modeling and Simulation Definitions RATIONALE This document establishes basic terms and definitions applicable for modeling and simulation of aircraft electric power systems. It lays the foundation for a series of AIRs and ARPs within the scope of the SAE AE-7M committee. INTRODUCTION B
6、ackground: The development of more- and all-electric aircraft (MEA, AEA) concepts for future-generation commercial and military airplanes has significantly impacted aircraft electric power system (EPS) design. Many functions that are conventionally managed by hydraulic, pneumatic and mechanical powe
7、r are being replaced by electric power aiming to reduce overall aircraft weight and size, improve fuel efficiency and reduce environmental impact. On the other hand, this transition has resulted in a substantial increase in the number of onboard electrical loads typically driven by power electronic
8、converters (PEC) and motor drives, e.g., for such functions as pumping fuel, cabin pressurization and conditioning, engine start, flight controls, landing gear actuation, and many others. Developing the EPS architectural bus paradigm for the next generation of aircrafts (more- and all-electric) invo
9、lves extensive modeling and simulation (M however they are interrelated and need to be considered as a product set. The HLA is an integrated approach that has been developed to provide a common architecture for simulation. 2.2.3 ARP4754 Rev. A “Guidelines for Development of Civil Aircraft and System
10、s“ ARP4754 addresses the process of aircraft systems development taking into account the overall aircraft operating environment and functions. This includes validation of requirements and verification of the design implementation for certification and product assurance. SAE INTERNATIONAL AIR6326 Pag
11、e 5 of 19 2.2.4 EIA Standard EIA-632 “Processes for Engineering a System“ EIA-632 describes ways of engineering and producing quality systems and focuses on conceptualizing, creating and realizing a system and the products that constitute it. The standard is applicable across all industry sectors an
12、d technology domains, describing the essential features of the engineering practices. 2.2.5 MATHWORKS Automotive Advisory Board (MAAB) “Control Algorithm Modeling Guidelines Using MATLAB, Simulinkand Stateflow“ In the subject document, MAAB sets guidelines for using Mathworks products to achieve suc
13、cessful M the same applies when representing properties such as hysteresis. These can be manually reviewed against the requirements for the particular modeling task. SENSITIVITY OF SOLUTION: In general terms, solution sensitivity is a model property (or characteristic) which evaluates the models rep
14、eatability, i.e., ability to deliver the same solution when running the model a (large) number of times under the same conditions. Well-built models should deliver reasonably close solutions. Good modeling practice requires evaluation of the confidence in the model (quality assurance). SENSITIVITY T
15、O PARAMETER AND INPUT CHANGES: These are used to describe how small variations or uncertainties in model parameters, or uncertainties in the definition of the system (mathematical model) inputs influence the solution of the system. These are, in general, characteristics of the studied phenomenon/sys
16、tem, in contrast to sensitivity of solution which is a characteristic of the model. STATE CONSISTENCY: The model is said to be state-consistent if all the state variables necessary to fulfill the requirements of the models intended use are represented within the model. EVENT PHASE ORDERING: This rel
17、ates to sequencing of modeled events: Will the model produce the required events or state changes in the correct order, in relation to outside signals and/or in relation to each other? REAL-TIME MODEL: The model is said to be a real-time model if it can execute as fast or faster than wall-clock time
18、, not just on average, but at each and every time step. Typically, real-time models run at fixed time steps. SAE INTERNATIONAL AIR6326 Page 9 of 19 SOFTWARE-IN-THE-LOOP: Software-in-the-Loop (SIL) is a simulation technique that is used for the development of complex real-time systems. The plant mode
19、l is simulated using an appropriate solver, while the controller model typically uses a discrete time step, is compiled and runs on a computer (as opposed to a hardware target or emulator). This technique allows the user to evaluate the software execution under different input conditions. SIL M henc
20、e, all its complexity is included when testing the control platform. The controller is cross-compiled and runs on the target hardware. EXPERIMENTAL HARDWARE-IN-THE-LOOP: Experimental Hardware-in-the-Loop (xHIL) is a technique that focuses on the integration of multiple system components. Both the pl
21、ant and the controller are physical hardware with emulated and/or simulated subsystems supplying boundary conditions to each other. Parts of the system may be emulated. For example, an engine might be emulated using a drive stand in order to perform integrated testing of a generator. 3. DEFINITION O
22、F MODELING LEVELS This section defines the multi-level paradigm for aircraft EPS M hence, it may be used for addressing EMC phenomena. SAE INTERNATIONAL AIR6326 Page 10 of 19 The modeling boundary for this level does not normally extend beyond the device itself. The device environment is modeled as
23、a set of simplified boundary conditions. Hence, this level is of very little or no practical use for simulation of entire EPS architectures or in sections sufficient for analysis of source(s)-to-load(s) studies. ARCHITECTURAL LEVEL FUNCTIONAL LEVEL BEHAVIORAL LEVEL DEVICE PHYSICAL LEVEL Figure 1 - M
24、ulti-level modeling paradigm 3.2 Behavioral Level The next level is the behavioral level, which uses lumped parameter subsystem models and targets frequencies up to hundreds of kHz. Behavioral models have enough fidelity to include power electronic converter switching (idealized or deviced-based) an
25、d to study the impact of harmonics and conducted EMC. Models at the behavioral level generate signals representative of actual hardware waveforms and thus can address, e.g., detailed harmonic analysis, design of passive and active filters, and conducted EMC phenomena. The modeled system for this lev
26、el is normally a subsystem that may extend to several power converters. The EPS can be modeled as a set of boundary variables (e.g., representative voltage), current sources and/or representative impedances. Typically, this level would be the most detailed one for use in EPS-level studies. Example o
27、f behavioral modeling is given below. 3.3 Functional Level Above the behavioral level is the functional level, which addresses low-frequency transient behavior. Here the guideline is that the power system components are modeled to handle dynamic frequencies up to 100 to 150 Hz (which corresponds to
28、one third of base grid frequency for a typical 400 Hz aircraft EPS), with waveform accuracy compared to the behavioral model within a few percent. Hence, the prime purpose of this level includes simulation-based studies of EPS dynamics and stability, response to load impacts and shedding, and low-fr
29、equency power quality. This level utilizes non-switching or averaged PEC models and allows modeling the EPS either in its entirety or in sections sufficiently large to obtain a holistic generator-to-load dynamic overview. The models developed at this level are suitable for linearization in order to
30、undertake classical small-signal analysis. The functional level may be conditionally subdivided into two sublevels, with a second or “extended functional” level. The extended functional level includes such effects as ripple and dominant harmonics, detailed control structures and physical coupling wi
31、th thermal, mechanical and other domains. Ripple, harmonic or switching frequencies can be incorporated through advanced baseband transformation techniques that avoid the need to time simulate through power converter switching events. SAE INTERNATIONAL AIR6326 Page 11 of 19 3.4 Architectural Level T
32、he top architectural level, as its name implies, aims to facilitate top-level, global EPS architecture studies. The architectural level does not model transient dynamics, but considers the global system in steady state. The architectural level may be conditionally separated into two sub-levels accor
33、ding to the tasks addressed as follows: In the logical sub-level, the EPS architecture is studied for its functional integrity. The models consist of boolean states and control logic laws that allow fault impact studies, component disconnection and bus (re-)configuration. This level may be used for
34、reliability studies when models are characterized with failure probability constants. The requirements sub-level targets the sizing of EPS components to match the set of given load specifications. The level studies nominal, or steady-state, power flow and thus facilitates the sizing of EPS component
35、s (converters, generators, cabling, contactors, etc.) with respect to rating, weight, and cost across the range of flight scenarios. The requirement level should also include event switching and reconfiguration modeling since the steady state power flow is affected by reconfigurations. Although the
36、architectural sub-level does not represent EPS dynamics (for example, control loops are idealized to their referenced values), under exceptional circumstances the level may include significant dynamic effects approximated to first-order responses. A summary of the preceding modeling levels, includin
37、g basic characteristics and typical modeling tasks for that level, is given in Table 1. Table 1 - Modeling levels and their basic characteristics Model Level Type of Model Example Studies ARCHITECTURAL - Boolean states - Failure probability constants - Control logic - Steady state solutions - Ideali
38、zed control loops o EPS architecture functional integrity o Logic control law verification o Preliminary EPS sizing (weight, cost, power flow and cabling arrangements, power management) o EPS events and reconfiguration modeling FUNCTIONAL - State average models - Non-switching, but (possibly) repres
39、entative of ripple and higher harmonics - System and subsystem control structure - Detailed control structure o Fine EPS sizing o Basic EPS dynamics (up to 1/3 of grid frequency): turn-on /turn-off transients, EPS stability, fault development and clearance, energy management functions o Low-frequenc
40、y power quality o Generator-to-load dynamic overview o Coupled electrical, thermal, and mechanical effects o Multiple frequency selection through advanced base-band transformation modeling techniques BEHAVIORAL - Idealized switches - No high-level control loops o Detailed power quality (harmonic ana
41、lysis and conducted EMC) o Active and passive filtering o Fault protection studies o Global correlation with tests DEVICE PHYSICAL - Physics-based switches - Device-scale parasitic effects - Multi-physics coupling o Detailed component design including mechanical and thermal behavior o Conducted and
42、radiated EMC o Verification and in-depth analysis o Local correlation with tests SAE INTERNATIONAL AIR6326 Page 12 of 19 4. EXAMPLE: MEA EPS MODELING AT THE FUNCTIONAL AND BEHAVIORAL LEVELS This section demonstrates the application of the multi-level modeling paradigm to an MEA EPS simulation study.
43、 The example EPS shown in Figure 2 represents the hybrid more-open electric technologies (MOET)-type EPS that is discussed in detail in the paper referenced in Reference 2 in 2.1; here it is used to demonstrate the application of functional and behavioral levels of EPS modeling for investigation of
44、system-level events in both normal operation (transients in response of ECS start-up and actuator operation, including regenerative regime) and in abnormal regimes (loss of one generator, line-to-line fault, HVDC bus fault). Figure 2 - Hybrid EPS architecture 4.1 Hybrid MEA EPS The EPS is supplied b
45、y two 125 kVA synchronous generators (SG) whose output voltages are regulated by generator control units (GCU). The main ac buses, HVAC1 and HVAC2, are 230 V, 360 to 900 Hz frequency-wild. An autotransformer-rectifier unit (ATRU), ATRU1, rated at 60 kVA converts HVAC1 ac voltage into dc voltage to f
46、eed a 540 V dc bus, HVDC1. A 30 kW environmental control system, ECS1, based on a permanent-magnet machine controlled through a pulse-width modulated (PWM) inverter unit, CIU, is connected to HVDC1 together with other dc loads. HVAC1 is also feeding the WIPS and the essential ac bus, AC ESS. All fli
47、ght critical equipment is fed from this bus, therefore in the real system this bus is supported with batteries through dedicated power-electronic converters. In this example we ignore battery support since this does not influence the EPS dynamics; instead we consider two electromechanical actuators,
48、 EMA1 and EMA2, fed from the AC ESS. Both EMAs (rated at 5 kW each) are based on permanent-magnet machine drive systems with back-to-back PWM converters. Similar to HVAC1, under normal conditions HVAC2 feeds ATRU2 to create HVDC2 with ECS2 and other dc loads. There is no WIPS for HVAC2, however ther
49、e are a few additional ac loads including an autotransformer to create single-phase 115 V for the legacy loads. SAE INTERNATIONAL AIR6326 Page 13 of 19 The functional-level modeling for this example is implemented employing dq0-approach as reported in 0,0; other approaches are also possible. 4.2 Normal Operation Scenario Studies Under normal operating conditions the EPS consists of two “islands” associated with HVAC1 and HVAC2.
