Tracking-protection-recovery switching control for aero-engines

https://doi.org/10.1016/j.jfranklin.2017.09.028Get rights and content

Abstract

This paper presents a novel tracking-protection-recovery switching strategy to solve the thrust tracking and safety protection multi-objective control problem for the aero-engines. The proposed switching control strategy overcomes the contradiction between the tracking performance and the safety requirement. The design procedure is with larger degree of freedom and less conservatism. The proposed switching controller can be designed in three steps. For the tracking stage, the tracking controller is designed only according to the rapidity requirement for the thrust tracking with less consideration of safety. For the protection stage, the protection controller is activated to limit the protected output in the safety region. Because of the properly designed protection controller, it is unnecessary to switch on the protection controller before the protected output reaches the safety boundary. That reduces conservatism and makes the tracking performance improved. For the recovery stage, the recovery controller, as well as the properly designed resetting law, is utilized to guarantee finite number of switches and the resulting asymptotic tracking. Because of the properly designed switch-off condition for the protection controller, the thrust tracking performance gets improved. The protected output is also successfully limited. Finally, a case study for a two-spool turbofan engine is performed to verify the effectiveness of the proposed scheme. It is also indicated that the proposed tracking-protection-recovery switching strategy can improve both safety performance and the tracking transient performance.

Introduction

Since engines are one of the most essential components of aircraft, research on engine control has attracted much attention in recent years [1], [2], [3]. Modeling and control of the aero-engine has become more and more important research field. Advanced control methods have been applied to the aero-engine control, such as adaptive control [4], distributed control [5], fuzzy control [6], gain scheduling control [7], linear parameter-varying (LPV) control [8], model predictive control (MPC) [9], sliding mode control (SMC) [10], and so on. The strong couplings among flight dynamics, aerodynamics, propulsion and control [1], [2], [3] result in various engine safety boundaries during the working process of the aero-craft. Safety boundaries usually include the combustion stabilization boundary, the combustor wall temperature limitation, the inlet channel un-start boundary, and so on. Disastrous accidents may happen once the system trajectory exceeds any one of the safety boundaries. Therefore, protection measures must be taken. On the other hand, to pursue desirable flight performance, the aero-engines are usually required to work close to the safety boundaries. For example, in work [23] and [24] controllers are designed to pursue performance such as fast response, handling quality and robust stability. The traditional safety protection measures usually sacrifice the rapidity performance to guarantee safety. One of the effective strategies to deal with the contradiction between the rapidity performance and safety requirements is to apply the switching control scheme [11], [12] to the multi-objective control problem.

Applying the switching multi-objective control scheme, the design procedure is with larger degree of freedom and less conservatism. The tracking controller and protection controllers can be designed separately. That reduces the complexity and enlarges the degree of freedom of design. Otherwise, it is more complicated if we consider the safety constraints simultaneously when designing the thrust tracking controller for an aero-engine. The properly designed switching law can select the intervals when protection controllers are necessary, leaving more intervals for the system to pursue tracking performance. The multi-objective switching control strategy is expected to bring improvement on both thrust tracking performance and safety.

Several results on multi-objective switching control are available in the literature [13], [14], [15], [16], [17], [18], [19], [20], [38]. The widely used min/max control switching law is studied in work [13] and [14]. However, the safety performance, finite number of switching and stability are all not guaranteed theoretically. The min/max control switching scheme is combined with sliding mode control to ensure safety in [10]. The result in [10] is extended to the multi-input case to make full use of the control inputs [39]. However, a properly designed switch-off condition may bring more freedom to pursue tracking performance. Moreover, an associated sliding mode surface is designed for the fan speed set-point objective, which may also limit the system to pursue tracking performance.

To solve the multi-objective control problem using switching strategy, we must guarantee transient performance as much as possible in the safety region. An expected switching law, including switch-on condition and switch-off condition for the protection controller, should be properly designed to guarantee both thrust control performance and safety. A safety-margin-dependent switching law was proposed [15], [16], [17], [18]. The protection controller is activated at precisely necessary moments. It is reasonable that more information than only the control input must be used. But the switching designed only according to the safety margin may bring unexpected control discontinuity and thus deteriorate the performance. The work [19] designs a bumpless switching controller to reduce this control discontinuity. A coordinated switching strategy is presented to take advantage of the two switching methods above in [20]. But no stability analysis is included.

However, how to conduct the dual design of the controller and the switching law to improve both thrust performance and safety margin is an essential problem. The well-developed switched system theory [25], [26], [27], [28], [29], [30], [31], [32], [37] provides no methods to solve this problem. This motivates our research work.

In this work, a novel tracking-protection-recovery switching strategy will be proposed to solve the thrust tracking and safety protection multi-objective control problem for the aero-engines. The contradiction between the tracking performance and the safety requirement will be largely overcome by the proposed strategy, the one with larger design degree of freedom and less conservatism. The switching controller can be designed in three steps. For the tracking stage, the tracking controller will be designed to make the fan speed track the command as quickly as possible. For the protection stage, the protection controller will be activated as the system trajectory reaches the safety boundary to ensure safety. For the recovery stage, the recovery controller and the resetting law will be properly designed to ensure invariance and stability. Finally, a case study for a two-spool turbofan engine model will be performed to verify the effectiveness of the given method.

Section snippets

Multi-objective control problem arising from aero-engine control

For an aircraft, an aero-engine control system with high performance is required to achieve high performance flight. But there are many limitations during the working process. High performance is usually achieved very near the safety limitations. Control design must make a compromise between performance and in Fig. 1, the safe working region is shown. The fan speed, as an important representation of thrust, is usually selected as the controlled output. During an acceleration procedure, for

Main results

In this section, we will propose a tracking-protection-recovery multi-objective switching control strategy. An asymptotic tracking condition for the closed-loop switched system will be given.

Case study

We will verify the effectiveness of the proposed tracking-protection-recovery switching strategy through the two-spool turbofan engine model in work [10]. The considered simplified model can approximate the aero-engine well under the step command with small amplitude such as the step signal yr1(t) we consider here. The parameters are the same as in the motivating example. Obviously, the given parameters satisfy the Assumption1 to Assumption3.

For the tracking stage, the tracking PI control

Conclusion

A novel tracking-protection-recovery multi-objective switching strategy has been proposed in this paper. As a solution to the thrust tracking and safety protection control problem for the aero-engines, the proposed strategy overcomes the contradiction between the tracking performance and the safety requirement, with larger design degree of freedom and less conservatism. The proposed switching controller can be designed in three steps.

For the tracking stage, we take advantage of the

References (40)

  • M. Pakmehr et al.

    Distributed control of turbofan engines

  • A.J. Chipperfield et al.

    Fuzzy scheduling control of a gas turbine aero-engine: a multiobjective approach

    IEEE Trans. Ind. Electron.

    (2002)
  • M. Pakmehr et al.

    Gain scheduled control of gas turbine engines: stability and verification

    ASME J. Eng. Gas Turbines Power

    (2014)
  • G. Balas

    Linear parameter-varying control and its application to a turbofan engine

    Int. J. Robust Nonlinear Control

    (2002)
  • H. Richter et al.

    Multiplexed predictive control of a large commercial turbofan engine

    AIAA, J. Guid. Control, Dyn.

    (2008)
  • D.R. Yu et al.

    Multi-objective robust regulating and protecting control for aero engines

    J. Eng. Gas Turbines Power

    (2008)
  • X. Liu Multi-objective regulating and protecting control for aero-engines, Ph.D. dissertation, Harbin Institute of...
  • C. Chao et al.

    Safety protection switching control for aero-engines based on multiple Lyapunov functions method

  • C. Chen et al.

    Safety protection control for a class of linear systems via switching strategy

  • C. Chen et al.

    Command switching strategy based safety protection control for aeroengines

  • Cited by (0)

    This work was supported by the National Natural Science Foundation of China under Grants 61773098 and 61773100 and the 111 Project (B16009).

    View full text