Meta-heuristic global optimization algorithms for aircraft engines modelling and controller design; A review, research challenges, and exploring the future

TLDR: The potential solutions for the remaining research challenges for application to aerospace engineering problems in the future include the implementation of enhanced and modified optimization algorithms and hybrid optimization algorithms in order to achieve optimal results for the advanced engine modelling and controller design procedure with affordable computational effort.

About: This article is published in Progress in Aerospace Sciences.The article was published on 2019-01-01 and is currently open access. It has received 38 citations till now. The article focuses on the topics: Global optimization & Genetic algorithm.

Figures

Table 2 – Objective function samples for controller gain tuning

Figure 3. The schematic and characteristics of the modelled turbojet engine (images from [37])

Figure 9. Min-Max control strategy for a single spool turbojet engine

Table 1 – Comparison between different categories of global optimization algorithms

Figure 2. The schematic of NGDF for modelling jet engines [34]

Figure 13 – Dynamic convergence comparison of optimization algorithms for MinMax controller gain tuning of the turbojet engine (average of 15 runs)

Frequently asked questions

Q1. What are the contributions in "Ea: evolutionary algorithms epr: engine pressure ratio go: global optimization gte: gas turbine engine hps: high pressure shaft ifpc: integrated flight/propulsion control lps: low pressure shaft lp: linear programming mea: more electric aircraft mgo: meta-heuristics global optimization milp: mixed-integer linear programming problems minlp: mixed-integer non-linear programming problems" ?

To deal with these challenges, two efficient optimization algorithms, Competent Genetic Algorithm in single objective feature and aggregative gradient-based algorithm in multi-objective feature are proposed and applied in a turbojet engine controller gaintuning problem as a case study. Based on this comparison and analysis, the potential solutions for the remaining research challenges for application to aerospace engineering problems in the future include the implementation of enhanced and modified optimization algorithms and hybrid optimization algorithms in order to achieve optimal results for the advanced engine modelling and controller design procedure with affordable computational effort. 

Q2. What have the authors stated for future works in "Ea: evolutionary algorithms epr: engine pressure ratio go: global optimization gte: gas turbine engine hps: high pressure shaft ifpc: integrated flight/propulsion control lps: low pressure shaft lp: linear programming mea: more electric aircraft mgo: meta-heuristics global optimization milp: mixed-integer linear programming problems minlp: mixed-integer non-linear programming problems" ?

In order to explore the future of the MGO methods in GTE applications it is worthy to summarize the current ongoing progresses in the GTE field. Therefore, new objectives and concerns in the future controllers should be considered which are clearly more sophisticated and demanding than the current ones. This matter causes compromising effects on the objective function definition of the engineering problems [ 83-85 ]. Table 5 summarizes these challenges and potential solution. 

Q3. What are the main objectives of the proposed research challenges for MGOs in GTE?

To use advanced computational methods (e.g. linkage learning techniques) toenhance the efficiency of the MGOs and to increase their affordability for dealing with huge problems; • 

Q4. What are the different classes of deterministic global optimization techniques?

Linear Programming (LP) [6], Mixed-integer linear programming problems (MILP) [7], Non-linear programming problems (NLP) [8], and Mixed-integer non-linear programming problems (MINLP) [9] are different classes of deterministic global optimization techniques. 

Q5. What is the other approach for dealing with the research challenges of MGO techniques in GTE applications?

33The other approach for dealing with the research challenges of MGO techniques in GTE applications is to work on developing new types of algorithms by combination of MGO approaches and gradient-based methods [77] usually called “Hybrid Algorithm” as discussed in [53-54]. 

Q6. What is the effect of the second order term of the LPS on the engine?

In addition, Andoga et al. did a comprehensive literature survey on the reduced order model generation procedure for the gas turbine aero-engines in 2008 and with several simulations showed that the effect of the second order term of the LPS is not negligible in high bypass turbofan engines [36]. 

Q7. What is the schematic of the NGDF model?

The schematic of these models is presented in figure 1.14between the NGDF model and the models proposed by other researchers is that in NGDF the transfer functions between different inputs and outputs has an incremental form to enhance the accuracy of the model. 

Q8. What are the main progresses in the modeling and control of gas turbine engines?

Main progresses in modelling and control of gas turbine engines could be explained as follows:• NewGTEmodels [82]: the new, advanced models for gas turbines include morecomplicated details and parameters than previous ones.